You are on page 1of 39

Group-Buying On The Web: A Comparison Of Price Discovery

Mechanisms


Krishnan S. Anand
1
and Ravi Aron
2

OPIM department, The Wharton School
University of Pennsylvania.





Abstract: Web-based Group-Buying mechanisms, a refinement of quantity
discounting, are being used for both Business-to-Business (B2B) and Business-to-
Consumer (B2C) transactions. In this paper, we survey currently operational online
Group-Buying markets, and then study this phenomenon using analytical models. We
surveyed over fifty active Group-Buying sites, and provide a comprehensive review of
Group-Buying practices in the B2B, B2C and non-profit sectors, across three continents.

On the modeling side, we build on the coordination literature in Information Economics
and the quantity-discounts literature in Operations to develop an analytical model of a
monopolist who uses web-based Group-Buying mechanisms under different kinds of
demand uncertainty. We derive the monopolist's optimal Group-Buying schedule, and
compare his profits with those that obtain under the more conventional posted-price
mechanism. We also study the effect of heterogeneity in the demand regimes, in
combination with uncertainty, on the relative performance of the two mechanisms. We
further study the impact of the timing of the pricing decision (vis-à-vis the production
decision) by modeling it as a two-stage game between the monopolist and buyers.
Finally, we investigate how Group-Buying schemes compare with posted price markets
when buyers can revise their prior valuation of products based on information received
from third parties (infomediaries).

In all cases, we characterize the conditions under which one mechanism outperforms the
other, and those under which the posted price and Group-Buy mechanisms lead to
identical seller revenues. Our results have implications for firms' choice of price
discovery mechanisms in electronic markets and scheduling of production and pricing
decisions in the presence (and absence) of scale economies of production.

1
Email: anandk@wharton.upenn.edu; Tel: (215) 898 1175.
2
Email: raviaron@wharton.upenn.edu; Tel:(215) 573 5677.

1

1. Introduction

Group-Buying schemes have been in vogue for very many years, particularly in the
context of selling on television via the popular Home Shopping Network. Web-based
variants of Group-Buying have received a lot of attention recently as part of the wave of
innovative online market-based mechanisms such as auctions, reverse auctions and
Priceline’s ‘name- your-own-price’ scheme. Nevertheless, unlike in the case of auctions
(which has a rich history of analytical research over more than forty years), Group-
Buying sorely lacks a theoretical framework.

This paper develops analytical models of Group-Buying, building on extant
academic research in the disciplines of Operations Management, Economics and
Information systems. As we demonstrate, two streams of research — the quantity-
discounts literature in Operations Management and the literature on price discovery under
demand uncertainty in Information Economics — serve as building blocks to develop a
theory of Group-Buying. We compare the performance of Group-Buying with that of the
market mechanism most widely used in electronic markets — simple posted prices —
and identify the conditions under which each dominates.

The remainder of this paper is organized as follows. In the next Section, we lay
out our research objectives. In Section 3, we provide a comprehensive review of the
current state of praxis, by surveying a variety of Group-Buying practices, world-wide, in
the B2B, B2C and non-profit sectors. In Section 4, we review the relevant prior research
and comment on important results that are relevant to our research. In Sections 5, 6 and 7
we develop three different analytical models of demand uncertainty and compare the
performance of the two market mechanisms (Group-Buying and posted pricing) in each
case. In Section 8, we discuss the managerial implications of our research and highlight
possible extensions.

Section 2: Research Objectives

Our objective is to study the phenomenon, widely seen in current practice, of web-based
Group-Buying. This is a specific instance of the proliferation of complex price-discovery
mechanisms enabled by the Internet and the World-Wide-Web. Electronic Group-Buying
market mechanisms seek to aggregate a variety of disparate buyers via the web—
remotely and asynchronously— essentially by providing them price-based incentives to
participate in the market and buy the good. In what is a classic volume-discounting
scheme, the market operator offers buyers lower prices for higher volumes procured
collectively. The Group-Buying markets that we surveyed were all characterized by
prices declining in sales volumes. Further, web-based Group-Buying was almost entirely
asynchronous— users have to work against some deadlines imposed by the seller, but
have considerable leeway to make their bids via the internet within the week(s) during
which the object is on sale. We provide a detailed exposition of several kinds of
electronic Group-Buying markets in Section 3.

2
We know that simple posted prices are widely used in consumer sales—in traditional
approaches such as retail and catalog selling, for instance, but also in web-based selling.
While the web facilitates complex selling mechanisms such as Group-Buying, a larger
question is how effective are such complex schemes, relative to simple posted prices, and
more importantly, under what conditions do they add the most value? Our study is
focused on answering two questions, both of immediate concern to current practice:

(i) Given that a firm uses the Group-Buying (volume discount ing) market mechanism to
sell its product(s), what is its optimal Group-Buying (price-quantity) schedule, and what
are its revenues under this schedule?

(ii) How does this optimal Group-Buying schedule compare with the other widely used
market mechanism, specifically, simple posted prices? Specifically, what market
conditions and product characteristics would justify the use of Group-Buying over posted
prices?

The practical application of insights into these two questions is obvious. To the
best of our knowledge, our paper is the first in academia to (i) identify and delineate the
Group-Buying market mechanism, (ii) develop analytical models of this mechanism, (iii)
study the optimal such schedule under a variety of parameterized market scenarios, and
(iv) analyze its performance relative to the simpler mechanism of posted prices (list
prices).

3. Group-Buying: The Current State of Praxis

We quote the mission statements of two market operators that we surveyed to illustrate
that volume discounting and demand aggregation are at the core of Group-Buying.

“Co-buying is co-operative shopping for the 21st century. It's a really simple way of getting
better value by bringing people together via the Internet. By bringing together as many members
as possible, LetsBuyIt.com can negotiate lower prices with merchant partners (also referred to as
suppliers) or manufacturers. The more people, the lower the prices.”
- Letsbuyit.com

“The more we are, the bigger the negotiation power towards trusted suppliers. The
immediate result? Prices that fall without having to negotiate on your own. The more we
are ("many"), the happier we are ("happy") because everyone pays less and everyone
benefits from a better service.”
- The HappyMany

In our study of currently operational
3
online markets, we found that Group-Buying
(referred to as ‘co-buying’ in the above quote) is a widely deployed price-discovery
mechanism in a variety of markets and contexts. These markets are not limited to a
particular geographic region (such as the United States) or a particular product category.
We will see that the Group-Buying mechanism is extensively used in the United States,

3
As of mid-September 2002.
3
Europe (including in Germany, France and the UK) and Asia (including in India,
Thailand, Singapore and Egypt). This review discusses representative, interesting
examples of this deployment. Given the large number of active Group-Buying sites (well
over 50), an exhaustive review is beyond the scope of this paper; however, a fairly
comprehensive list of 26 representative Group- Buying websites from the B2B, B2C and
non-profit sectors is provided in the Appendix at the end of this document. Group-Buying
schedules are employed for branded consumer products as well as intangible services
such as bandwidth and network security, in both business-to-business (B2B) and
business-to-consumer (B2C) markets, and in public as well as private markets. These
markets are characterized by the belief that both suppliers and buyers stand to benefit
through Group-buying. Evidence of this belief is adduced by the web-sites of both buyer
[IRPG, APPA, TBG, GBP]
4
and supplier [LBI, THM] consortia in a variety of markets,
that prominently tout the benefits of Group-Buying to their members. While we
documented markets in three continents that deploy the group-buying mechanism, we
have restricted the discussion below to some interesting instances that are representative
of several currently operational markets, or that employ theoretically interesting variants
of Group-Buying, of relevance to our modeling-based analysis.

The Group-Buying site e.conomy, operated by PricewaterhouseCoopers , is a successful
market based in the US [ECON]. This B2B market allows buyers to purchase indirect
goods and services such as office supplies, ‘temps’ (temporary staffing services),
furniture, commercial print, computer hardware, software, telecommunications &
connectivity, and company travel, at substantially reduced prices that result from higher
purchase volumes. Members are assured of an initial maximum price (a ‘price ceiling’).
As they place orders, incrementing the total volume demanded, the resulting price-
declines are broadcast (and available to all) until the market clears at a predetermined
time. Let’sBuyit.com, quoted in the beginning of this Section, is a consumer group-
buying market currently functional in the UK, Germany and France, that works in a
manner very similar to the e.conomy model. Demand is aggregated across multiple
buyers, and prices decline with increasing volumes until they reach a stable level. All
buyers pay the same market-clearing price. A nuance to the standard Group-Buying
mechanism that this company permits is that buyers can choose not to declare a price
ceiling—buyers’ greater risk without a price ceiling may be offset by guaranteed
availability of the good at the common, lowest price. In the words of Let’sBuyit.com,
“…Before you join a Co-buy, you need to decide whether you would like to buy at the
current price – the price reached when the Co-buy closes - or at the Best Price only. If
you choose to buy at the current price, you will receive the product (provided your
payment details are accepted) and you will pay the closing price. If you choose to buy at
the Best Price and the Best Price is not reached (i.e. the required number of participant
Co-buys is not reached), you will not receive the goods. Your order will be cancelled and
you will not be charged…” Product categories sold in this market include exercise
equipment, consumer electronics, sport and leisure, food and wine, and jewelry.


4
Capital letters in square parentheses refer to websites of companies listed in ‘Group-Buying Markets:
Citation List’ provided at the end of this document.
4
The Group-Buying Partnership in the United Kingdom is a consortium of small and
medium sized businesses which negotiates volume discounts from suppliers of items such
as electricity and gas, computer equipment, office supplies and vehicles, and passes these
discounts on to buyers [GBP]. They claim on their web-site that “Group Buying offers a
simple and cost effective opportunity for small to medium sized businesses to combine
their individual buying power for the overall benefit of each individual member” [GBP].
The mechanism works through a process of demand aggregation wherein buyers commit
to purchase quantities subject to certain price ceilings-- should the consortium succeed in
procuring the quantities at or below these price ceilings, the buyers are obliged to follow
through on their commitments. This is a direct correlate of purchasing commitment
conditional on demand realization modeled in our paper. A different flavor of this
practice can be found in the market called Chennai Online, which targets buyers who are
redistributors [COL]. This electronic group buying market (run out of India) is a private
market that allows buyers to place orders by specifying price-quantity schedules where
purchase is required only if the supplier is able to meet the price specified by the buyers.
What is interesting in this context is the order of events that lead to price discovery. In the
case of products such as software suites for office use, magazines and periodicals from
the popular press and computers (servers and PCs), the supplier announces prices at
different levels of demand (price-quantity schedules) and the buyers place bids serially
until the market clears at a particular price. In a second category of goods that includes
kitchen appliances, computer peripherals, consumer electronics and electrical appliances,
the sequence of demand solicitation and production (or procurement) of the goods is
reversed. In this case, suppliers commit to a group-buying schedule and demand is
realized when a stable price-quantity tuple is reached. Then, suppliers either produce the
goods (kitchen appliances and computer peripherals) or buy them from third parties
(consumer electronics and electrical appliances) in exact quantities. Thus, while
production and inventory stocking decisions precede pricing in the first category
(modeled in Case 1, Section 4 of our paper), the second category is an example of
production postponement (the Case 2, Section 4 model).

Another version of the Group-Buying mechanism called StockBuzz (run out of Thailand)
is operated by Asian producers of high quality yarn who sell their products to
manufacturers of high-end apparel [STBZ]. The apparel manufacturers forecast demand
based on a number of macroeconomic factors, past sales data, seasonality and, most
importantly, retailer feedback. Since most of the manufacturers sell to the same retailers
and in the same urban markets, their demand estimates tend to be highly correlated, and
reflect consumer demand cycles. The yarn producers in turn face uncertainty in demand,
which tends to fluctuate between a robust (high demand) regime and a weaker regime
reflective of an adverse consumer demand cycle. Thus the yarn producers estimate the
probabilities of these two market states, based on prior seasonal data and other available
information. Using the two estimated demand schedules as guide rails, they commit to a
schedule of prices for different quantities with prices declining in quantities. Buyers
(apparel manufacturers) are allowed to place conditional orders which they are then
obliged to honor should the price reach the levels specified in the order contract. In this
case, yarn suppliers use Group buying as a response to uncertainty about which of the
two states of demand will be realized (modeled in Cases 1 & 2, Section 2 of our paper).
5

Group-buying mechanisms are also observed in specialized markets. In the hotel and
restaurant industry the Independent Restaurant Purchasing Group (IRPG) brings
together over 3000 restaurants across the United States and negotiates volume-based
discounts with suppliers on their behalf [IRPG]. IRPG argues that the benefits of its
Group-Buying scheme, called the IRPG Volume Rebate Program, for its members are
two- fold. Firstly, buying power is enhanced for large chains as well as small restaurants:
“Our national and regional contracts are based on the total volume of our group versus
that of an individual restaurant. No matter how many restaurants you have the IRPG can
add strength and stability to your purchasing power.” Secondly, many manufacturers
insist on minimum volumes for doing business with them, out of efficiency
considerations arising from their production and transaction scale economies. Smaller
restaurants often find it difficult to meet these minimum criteria. As IRPG argues, “Most
restaurants, by themselves, would probably not meet the manufacturer's minimum
volumes. However, when combined with those of the entire membership they exceed the
minimum and in many instances qualify for further "multi- unit-allowances" normally
only available to large chain accounts.” [IRPG]. In the skin-care product market, the
market operator ‘The Buying Group’ [TBG] operates a relatively straightforward group-
buying scheme where buyers commit to quantities at a maximum price. The market
operator then negotiates a price from suppliers for the aggregate quantity, and shares the
benefits of the lower prices with the buyers (here again, we see ‘procurement
postponement’, studied in Section 4 of the paper.). A noteworthy feature seen in this
market is the pre-order feature wherein buyers ‘express an interest’ in a product that is
not currently featured in the market, and pay for their desired quantities upfront, at pre-
specified prices. If there is enough interest, the market operator (in this case, The Buying
Group) attempts to procure the product at a price that will make trade possible for at
least some buyers. If the market operator succeeds, he charges all buyers (who are willing
to pay the clearing price or more) the same floor price. If trade is not possible, the market
operator refunds the buyers’ paid- up amounts. The requirement that buyers commit to a
quantity and price by paying the amounts upfront ensures that buyers do not make
frivolous bids, and also provides an incentive for the market operator to negotiate with
suppliers.

Branded consumer goods are also sold via the Group-Buying mechanism. Two
companies, McNopoly and Online Choice, operate markets that allow consumers to bid
for specific branded items from a menu of offered items [MPLY, OLC]. If their suppliers
offer discounts that support a market-clearing price, those consumers whose bids for the
products were higher than the market-clearing price are required to buy the products; the
others exit the market.

Many firms operating in the non-profit sector are also using Group-Buying. We found
that the Maryland Public Service Commission supports individual buyers in their
efforts to form buying groups to negotiate better rates for higher aggregate levels of
power consumption [MPSC]. Similarly, the American Public Power Association and
the Environmental Action Foundation have launched a number of Group-buying
initiatives that bring together consumers to negotiate lower rates with power utilities
across the US [APPA].
6

To summarize, we find evidence of widespread deployment of Group-Buying as a price-
discovery mechanism in both B2B and B2C sectors. A key feature that sets all Group-
Buying schemes apart from other market mechanisms is suppliers’ beliefs that pre-
committing to a price-quantity schedule where the prices are monotonically declining in
total purchase quantities (and not just an individual buyer’s purchase quantities) will
maximize supplier revenues by inducing greater buyer demand. Thus, Group-Buying is
very often targeted towards buyers with low bargaining power-- individual consumers or
small to medium businesses. In fact, many websites across all types of products, services
and countries, explicitly and prominently claim that ‘enhanced buyer bargaining power’
is the single biggest advantage of Group-Buying. The market operator Let’sBuyIt.com
summarizes this belief: “We make it easy for our members to come together online to
benefit from collective purchasing power. The more members that join a Co-buy, the
lower the price becomes….” [LBI]. Similarly, the Group-Buying intermediary Printing
Industries of New England (PINE) asserts on its website, “PINE has made
arrangements with several companies — companies you use in running your business —
to provide discounts on their products and services. PINE combines the overall buying
power of its 513 member companies to negotiate these discounts. These are discounts you
may not be able to receive if you were to negotiate directly with each company” [PINE].
The construction industry’s portal site in India, BuildByte [BBYT] puts it even more
succinctly, “Bigger the volume - Lesser the price.”
We saw that the Group-Buying schemes also differ in the order of events that lead to
price discovery. We found that, in some cases, suppliers first commit to a price schedule
leading to demand realization, and then produce (or procure) the products in volumes
exactly calibrated to the realized demand. In other cases, suppliers first produce (or
procure) products and then sell via a Group-Buy market through a volume discounting
mechanism. Ceteris paribus, the former approach leads to higher supplier profits but is
not always feasible due to long production/procurement leadtimes. However, the
important practical question of whether to adopt Group-Buying when production
postponement is or is not feasible hinges on the relative performance of Group-Buying
vis-à-vis posted pricing in these two cases. Since the answer is by no means obvious, our
model and analysis in Section 5 attempts to shed light on this question.
More generally, the widespread use of Group-Buying in practice is predicated on a
number of unverified, seemingly ‘commonsensical’ assumptions (reflected above in the
quotes from various companies). Our paper aims to provide a more analytical response to
this discourse, and verifies the extent to which these underlying beliefs are true under
various market conditions, through the artifice of mathematical modeling.

3. The Group-Buying Mechanism: Theoretical Underpinnings

The Group-Buy business model has two components: (i) a quantity discount scheme
offered by the seller; and (ii) decentralized decision making by the buyers (consumers),
who need to coordinate their actions to the extent possible. Thus, two streams of literature
7
are particularly pertinent to our analysis: (i) Quantity Discounts; and (ii) the Coordination
Problem. These are individually discussed below.

3.1 Quantity Discounts

Quantity (or Volume) discounts have a long and well-studied history in the
context of trades between businesses. There is a rich stream of academic literature in
Operations on quantity discount schemes [cf. Monahan (1984), Lee et al. (1986), Kohli et
al. (1989)]. The purpose of quantity discounts is to encourage the buyer(s) to purchase
more of the seller’s good, or, at least, in larger batches. There are two popular forms of
quantity discounts—all-units and incremental [Nahmias (1997)]. The usual structure of
such schemes is that there are breakpoints defining changes (decreases) in the unit cost.
Under all-units, the discount is applied to all the units in a given order; under the
incremental discount scheme, the discount applies only to additional units beyond the
breakpoint (see Nahmias (1997) for examples of each). The more popular all- units
discount scheme is relevant to Group-Buy schemes employed by the Home Shopping
Network, MobShop and other players. Kohli et al. (1989) make the strong claim in their
paper that “from a transaction-efficiency perspective, the choice between incremental and
all-units discounts is a matter of firm or industry practice, not a result of their desirability
for the buyer, the seller, or the buyer-seller system.” As we will demonstrate in the paper,
this finding (which applies to quantity discounts in a traditional context) does not hold for
the case of Group-Buying, where the outcomes for both buyer and seller critically hinge
on the structure of the discounts offered. In fact, it is very important under the Group-
Buying scheme that early bidders are not penalized by higher prices. The guarantee that
early bidders, who are presumably those with higher expected utility from the good, will
pay the lowest price offered to any buyer, promotes their entering the market early. This
lowers prices and encourages other customers to follow suit. Thus, consumers face a
coordination problem.

The Economic-Order-Quantity (EOQ) formula gives the optimal order quantity
for a buyer who has scale economies in procurement or production; the EOQ trades off
set-up costs and holding costs. Monahan (1984) extends this idea by deriving the optimal
pricing schedule that a supplier should offer, given that the buyer subsequently optimizes
her profits by ordering appropriate quantities. This induces EOQ-like buyer behavior, or
increases the optimal order quantity over that without quantity discounts. While Monahan
(1984) assumes that the supplier produces in lot sizes that mimic the buyer’s orders, Lee
et al. (1986) extend the analysis to the case where the supplier can use a different lot size
from the buyer. This makes the supplier’s analysis more challenging, since both
production batch-size and frequency are now decision variables. The supplier offers a
quantity discount scheme that optimizes over two different effects: (i) the buyer’s actual
orders (as in Monahan (1984)), and (ii) the supplier’s actual production schedule, which
determine his production and inventory costs. Kohli et al. (1989) analyze quantity
discounts as the outcome of cooperation (through bargaining or negotiation) between
buyer and supplier, and study the praetor-efficient outcomes of different axiomatic
bargaining solutions due to Nash, Kalai and Smordinsky, and Eliashberg.

8
A common feature of the academic literature discussed above [Monahan (1984),
Lee et al. (1986), Kohli et al. (1989)] is their focus on quantity discount schemes as an
instrument to improve transaction efficiency. While this may be part of the proffered
rationale for Group-Buy schemes, and more generally for all web-based models that
discharge the function of demand aggregation, the major use in practice of Group-Buy
channels has been for price discrimination. Economists have analyzed the use of
nonlinear pricing such as quantity discounts for second-degree price discrimination
[Tirole (1989)]. Group-Buying involves both waiting and uncertainty on the part of the
buyers before they know the outcome (the market clearing price). Firstly, the sale has to
be kept open long enough to ensure that a sufficient number of buyers arrive at the site.
Secondly, the final clearing price depends on both the number of buyers who arrive on
the site and their willingness to pay for the product. If there is a strong negative
correlation between customer willingness-to-wait and their valuation of the good (as is
often the case), higher- valued (impatient) customers could be served through retail
outlets or web-based posted prices; lower- valued (patient but price-sensitive) customers
could be served through Group-Buying, since they are willing to incur the additional
delay that Group-Buying entails in return for a lower expected price. Many of the
Group-Buy sites offer products (e.g. electronic items) that are less than state-of-the-art,
at steep discounts, while the latest product is sold in more conventional ways, using
web-based posted prices, or traditional channels such as bricks-and- mortar retailers. This
reduces the cannibalization of the sales of high margin, high quality items by low margin
items. Even in the absence of such quality-based second-degree price discrimination,
Group-Buy sites could simply serve heterogeneous customers through multiple channels.
Available evidence from the business press on this subject support these hypotheses.

3.2 The coordination problem

As discussed previously, a distinguishing feature of the Group-Buy business model is that
the traditional quantity discount model is overlaid with a coordination problem among
consumers. The pure coordination problem occurs when all players optimize a common
objective function; in this case, the players are said to be a team (Marschak (1955);
Marschak and Radner (1972)). Clearly, the customers under Group-Buy are not operating
as a team; they maximize their expected individual surpluses from their purchases.
However, each consumer’s behavior (whether bidding or not bidding, and the timing of
the bid) affects other consumers’ surplus. Thus, the literature on externalities is of
relevance to the Group-Buy problem.
5
Under negative externalities, each incremental
customer inflicts an extra cost (or loss of benefit) on all the others. The classic example
of a negative externality is the presence of congestion effects that dampen system
performance: each additional customer increases the system congestion and inflicts
additional delay costs on all other consumers. In contrast to such negative externalities,
the coordination problem in the case of Group-Buy is more benign – a bidding consumer
induces a positive externality on all other consumers, since his bid can only benefit other
consumers by lowering the expected clearing price. Further, since the final clearing price

5
There is a vast literature on externalities, a discussion of which would take us too far afield from the
focus of this paper. The famous ‘public good problem’ of economics is an example of this, and is
discussed in most graduate textbooks in economics (cf. Mas-Colell et. al (1995)).
9
is the same for all consumers (analogous to the all-units quantity discount model
discussed previously, rather than the incremental discount model), the consumers’
incentives are ‘almost team- like’, i.e., almost perfectly aligned. Under certain conditions,
the consumer will bid for the good even above his reservation value, since his bid may
induce others to bid, causing the price to fall below his reservation value. Thus, in
expectation, the consumer may be better off by bidding for the good above his
reservation value. The supplier can and should build this factor into his discounting
schedule. The appropriate schedule can trigger an avalanche of bids due to positive
externalities
6
; this feature of the Group-Buy scheme distinguishes it from all other market
mechanisms, notably posted prices and auctions.

4. Models of Demand Uncertainty

A monopolist’s choice of pricing mechanism is often driven by the nature of the
demand uncertainty that he faces. Web-based Group-Buying schemes are predicated on
offering buyers discounts for higher volumes of purchases. As discussed earlier, in a
Group-Buy scheme the seller declares a price-quantity schedule, where the price drops as
quantity increases. Buyers arrive at the market and declare their intent to buy at the
current price point. As the quantity demanded exceeds a predetermined (and
preannounced) level, the price drops to the next lower price point. This in turn
encourages more buyers to bid. All buyers get the good at the settlement price, which is
the price point at which the price stabilizes. Clearly, the settlement price is not more than
(and usually less than) buyers’ bids.
When the seller knows the structure of the demand with certainty, he will not resort
to a Group-Buying scheme; he can do as well by offering a fixed price that maximizes his
expected revenue. This is because, when the nature of the demand (function) is known,
the seller can calculate the exact settlement price from a Group-Buying scheme, and
simply announce a posted price equal to that settlement price. It is only when the seller is
not sure about the structure of the demand that he would resort to Group-Buy schemes.
We analyze the impact of different kinds of demand uncertainty on the seller’s pricing
strategy and compare the performance of the two pricing mechanisms (Group-Buying
and posted prices) under each.

4.1 Model of Parallel Demand Regimes

We construct a theoretical model of a monopolist-seller facing demand uncertainty.
The seller has an estimate of the demand for his product (at every possible price) but does
not know it exactly. The monopolist operates in one of two demand regimes (which we
term high / low) each of which is equally likely. The monopolist cannot observe or infer
the demand regime prior to the pricing decision. The market consists of a fixed number of
buyers transacting within a single period. Each buyer demands exactly one unit of an
indivisible good. The monopolist seeks to sell a quantity Q. The demand for the

6
The term ‘externalities’ here should not be confused with ‘network externalities’, which is a demand
side phenomenon under which the installed base of users of a network experience gains in value when
the size of the network expands due to adoption by new buyers. Network externalities are just one of
many different kinds of externalities.
10
monopolist’s product is determined by a number of factors, including product attributes
and consumer preferences, the availability of complementary or substitute products and a
host of macroeconomic factors. He has to choose from one of two pricing mechanisms:
Group-Buying or posted prices.

As discussed above, a general feature of all Group-Buy schemes is that the price is
decreasing
7
in the total quantity demanded. We preserve this feature in our model, where
the price schedule consists of two or more price points corresponding to different
quantities demanded. Further, the offered price
8
should be decreasing in the quantity
ordered, or equivalently, we need that
( ) ( )
, and , : .
i i j j j i j i
q p q p q q p p ∀ > ⇒ ≤

With the above formulation we turn to the monopolist’s revenue maximization problem
under Group-Buying.

We assume that demand for the good is linear, and given by: ( ) . Q p a m p · − ⋅
9
Under
the high and low demand regimes, the demands are given by:
( ) ( )
' ' ' '
and respectively, where .
h h l l h l
Q p a m p Q p a m p a a · − ⋅ · − ⋅ > Without loss of
generality, we normalize m to 1 and the resulting demand curves are given by:
and .
h l
q a p q a p · − · −
This formulation results in parallel demand curves for the two regimes. We will consider
the case when the two regimes have non-parallel demand curves in models that follow.
The two demand regimes are shown in Figure 1 below: the high demand regime is the
outer demand curve.

[INSERT FIGURE 1 HERE]

Thus the monopolist’s problem is to pick a price-quantity schedule that will
maximize his expected total revenue
10
without knowing which of the two demand
regimes is realized. Since there are only two demand regimes, the monopolist will pick a
pair of price-quantity tuples ( ) ( )
1 1 2 2
[ , , , ] q p q p , with one tuple corresponding to each
regime, such that the expected revenue is maximized, subject to the condition that
2 1 2 1
and q q p p > ≤ . The notation ( ) ( )
1 1 2 2
[ , , , ] q p q p means that the seller charges the
price
1
p when the total sales quantity is
1
q ≤ , and
2
p when the total sales quantity is
1
q >

7
We do not assume monotonicity although monotonicity would strengthen our results.
8
We distinguish between offered and realized prices here since a single price may be realized which is
one of the many offered prices.
9
The linear demand curve is of course widely used in the Economics literature, to capture the
relationship between prices and quantities (cf. Tirole (1989)).
10
Since production precedes pricing in our model, we are justified in treating the marginal cost as sunk
and therefore maximizing revenues (as opposed to revenues net of cost of production). It will be clear
from inspection that the solution and the ordering of market mechanisms by revenues will not change
if a positive marginal cost were to be factored into the analysis. When pricing precedes production, the
marginal cost might matter: in this case (studied in a later Section), our analysis assumes a non-zero
marginal cost.
11
but
2
q ≤ . Further, incentive compatibility constraints have to be satisfied, to ensure that
under each regime, the price-quantity pair that is operationalized is the one that the
monopolist intended for that regime. Thus, under each regime, the equilibrium outcome
should be that the customers, while maximizing their own value, prefer the price-quantity
pair intended for them, rather than the alternative. The implementation of this idea, in
connection with the well-known revelation principle of economics,
11
will become clearer
in the exposition below.

The seller has two alternatives open to him: (i) pick the higher price point from
the high demand regime (and therefore the lower price point from the lower demand
regime). This would imply that he expects to realize a higher quantity of sales from the
lower demand regime
12
; (ii) pick the lower price point (and therefore higher sales
quantity) from the higher demand regime and the higher price point (and therefore lower
sales quantity) from the low demand regime. Figures (2) and (3) illustrate options (i) and
(ii) respectively.

Case 1: Higher price point is picked from the high demand regime (see figure 2).

[INSERT FIGURE 2 HERE]

In Figure 2,
1 1 2 2
and
h l
q a p q a p · − · − . The quantity
1
q acts as a limit for the higher
price
1
p . The seller will offer the following pricing scheme
13
:

1 1 1
2
if quantity q ,
Price
otherwise,
h
p q a p
P
p
≤ · −
¹
·
'
¹

subject to the constraints
14

1 2 1 2
and . p p q q ≥ ≤ Under this scheme the price (outcome)
will always be
2
p irrespective of the demand regime that is realized. Consider the
following two scenarios: If the lower demand regime is realized, the price that will
prevail is
2
p and the quantity sold will be given by
2 2 l
q a p · − . If the higher demand
regime is realized, then seller would want the price to remain at
1
p ; however, buyers in
this demand curve would demand a quantity given by
2 2 h
q a p · − which is greater than
1 h
a p − . As a result, the seller would have to offer the lower price corresponding to the
higher quantity level. This leads us to our first result about the seller’s optimal revenue.

Lemma 1: The seller’s total revenue in expectation when he chooses the higher price
point from the higher demand regime is given by:


11
For an excellent exposition of the Revelation Principle, see Mas-Colell et. al (1995) or Myerson
(1991).
12
A higher price implies a lower quantity and vice versa in a Group-Buy pricing schedule.
13
It will be clear by inspection that this pricing scheme maximizes the seller’s revenues under Case 1.
14
These constraints result from the definition of Group-Buy mechanisms: higher quantities lead to
lower prices.
12

( ) ( )
( )
2
*
1 1 2 2
, , , ,
16
h l
a a
q p q p π
+
· 1
¸ ]
and the optimal price
*
P is given by
*
.
4
h l
a a
P
+
·
15
(3.1)


Remarks:

1. The above scheme is in essence a single price scheme (fixed price) since only a
single price will prevail irrespective of which demand is realized.
2. We do not specify what
1
p should be since this price is never realized. By
assumption of this Case, the only constraint on
1 1 2
is that p p p > .

Case 2: The seller picks the lower price point from the high demand regime and the
higher price point from the low demand regime.

[INSERT FIGURE 3 HERE]

In this case, the seller will offer the following pricing scheme:

1 1 2
2
if quantity q ,
Price
otherwise,
l
p q a p
P
p
≤ · −
¹
·
'
¹

subject to the constraints
1 2 1 2
and . p p q q ≥ ≤

Now the separator between the two pricing points is the quantity
'
2 q , which ensures that,
in the event the low demand regime is realized, the price does not slide down to
2
p . In
this case separation of prices across the two demand regimes seems possible.
16
Of
course, if the seller could maximize revenues individually in the two markets, the
revenue-maximizing prices in the high and low demand markets would be, respectively,
* *
1 2
and
2 2
l h
a a
p p · · , with the corresponding sales quantities
* *
1 2
and
2 2
l h
a a
q q · · .
However, this solution is infeasible under the Group-Buy mechanism, since the constraint
that the price should be decreasing in quantity demanded is violated (since
* * * *
1 2 1 2
p p q q > > and ). Since separate maximization of revenues across the two
demand regimes leads to an infeasible price-quantity schedule, we need to maximize the
seller’s revenue within the feasible set of prices. Lemma 2 provides the optimal solution.


15
Proof of this and all other propositions can be found in the additional document entitled “Proof Of
All Results”.
16
Recall that the seller’s objective is to implement a different price in each demand regime to optimize
his total expected revenue.
13
Lemma 2: The optimal (revenue-maximizing) prices and seller profits under group-
buying are given by
( ) ( )
( )
( )
( ) ( )
2
* *
1 2 1 1 2 2
2
* *
1 2 1 1 2 2
and , , , , if 2 1 ;
4 16
and , , , , otherwise.
2 8
l h l h
G l h
h h
G
a a a a
p p p q p q a a
a a
p p p q p q
π
π
¹
+ +
· · · ≥ − 1 ¹
¸ ]
¹
'
¹
· · · 1
¸ ] ¹
¹

In particular, any feasible group-buy schedule that sets the prices such that
1 2
p p
∗ ∗

yields revenues less than the above solution.

Remarks: Thus the monopolist is unable to exploit the Group-Buying price schedule to
implement discriminatory pricing based on which demand regime is realized. Since
* *
1 2
p p · under the optimal Group-Buy scheme (i.e., the price is independent of the
demand regime), the monopolist can clearly do as well using posted prices, by setting his
price
* * *
1 2
( ) p p p · · . The seller’s revenues under Group-Buying and posted prices are
identical.

The following Proposition derives from Lemmas 1 and 2.

Proposition 1: Under the above model of demand uncertainty, the seller’s revenues from
Group-Buying can never exceed his revenues from simple posted prices. In fact, the
optimal (revenue-maximizing) Group-Buying scheme simply mimics the optimal posted
price. Any deviation from posted pricing (such as differential pricing to discriminate
between realized demand regimes) will lead to sub-optimal revenues for the seller.

Remarks:

1. The above analysis shows that the revenue from Group-Buy schemes may at best
equal the revenue to the monopolist from posted pricing. However, when the two
revenues are equal it is important to note that the Group-Buy scheme does in
effect act as a posted price market. In Case 1, the price
1
p always slides down to
2
p , thereby establishing a single-price equilibrium across the two demand
regimes. In Case 2, the monopolist is able to offer a separation of prices based on
quantity demanded, but the revenues from posted prices dominate the resulting
revenues from Group-Buying.

4.2 Model of Intersecting Demand Regimes

In the previous section, we considered the case of demand uncertainty wherein the two
demand regimes were parallel. This implied that at any price the quantity demanded
under one demand regime dominated that demanded under the other. We now consider
demand uncertainty where one demand regime does not universally dominate the other.
Specifically, we study the case when the curves for the two demand regimes intersect:
one regime may result in a higher quantity demanded for one range of prices, while the
14
other regime may dominate over another range of prices (see Figure 4). We derive the
optimal revenues from Group-Buying and posted price markets and compare the two.

[INSERT FIGURE 4 HERE]

Figure 4 shows a case of intersecting demand curves. The generic linear demand curve is
given by Q A mp · − . Let the two curves be given by:
1 2 1 2
and a>b . q a m p q b m p where and m m · − · − ≥
1
2
2
Let . We normalize to 1
m
m m
m
·
so that the two demand curves are given by q a mp · − and q b p · − , where a>b and
1 m≥ . Further, as seen in Figure 4, the two curves intersect only when
a
b
m
> , and we
assume that this condition holds.

Each demand regime can occur with equal likelihood, and the seller has to set prices
before the demand regime is realized. Under Group-Buying, the seller picks a price-
quantity schedule as before. He offers prices of
1
p and
2
p for quantities demanded up to
the thresholds
1
q and
2
q respectively, where
1 2 1 2
and . q q p p ≤ ≥ We now determine the
maximum revenue and the optimal values of
1
p and
2
p under Group-Buying.

Lemma 3: For the model of intersecting demand regimes described above, the profit
maximizing Group-Buy price-quantity schedule is given by:

*
1
*
2
, if quantity demanded ;
2 2
Price *
, otherwise.
2
b a
q b
p
m
P
a
p
m
¹
≤ −
¹ ¸ _
¹
· ·
'

¹ ¸ ,
¹
¹
(3.2)
The seller’s expected revenue is
( )
2 2
* *
1 2
,
8
a mb
p p
m
π
+
· .
(3.3)


Remarks:

1. The seller chooses the quantity
2
a
b
m
− to separate the two prices. This should be
seen as a preventive measure to stop the market from clearing at a sub-optimal
price. This price-separation quantity can be explained as follows. The seller wants
to realize
1 1
( , ) p q when the realized regime is q = b - p, and
2 2
( , ) p q when the
realized regime is q = a - mp. If the demand regime given by b p − is realized,
the price could slide down to
2
p when the optimal price is the higher
1
p (see
figure 4). At a price of
2
p , the induced demand (under the regime of q = b - p ) is
15
given by
'
2 2
2
a
q b p b
m
· − · − . However when the seller specifies that the
minimum quantity demanded (for a price of
2
p to be offered) should be greater
than
2
a
b
m
− it prevents the market from clearing at a price of
2
p under this
demand regime.
2. If the demand regime given by a mp − is realized, then the choice of
'
2
2
a
q b
m
· −
as the quantity-discount threshold will not affect the market-clearing price since
the demand induced by
* *
2 2
is
2 2
a a
p q
m
· · , and we can easily check that
* '
2 2
2
a
q q b
m
> · − . Thus the choice of
'
2
2
a
q b
m
· − as the quantity-discount
threshold preserves the profit maximizing prices under both regimes for the seller.

We now derive the seller’s optimal price and revenues under posted pricing, and
compare these with the optimal prices and revenues under the Group-Buying
mechanism, in the following proposition.

Proposition 2: In the case of intersecting demand regimes,
(i) The profit maximizing price and revenues under posted pricing are given by
( )
( )
( )
( )
2
* *
and
2 1 8 1
a b
a b
p p
m m
π
+
+
· ·
+ +
.
(ii) The difference in revenues between the optimal schedules for Group-Buying
and posted pricing is given by
( )
( )
2
8 1
g p
a bm
m m
π π

− ·
+
> 0. Thus, the revenue to
the seller from Group-Buying strictly dominates the revenue from posted
pricing.

Thus, unlike the earlier case of parallel demand curves (wherein the two demand
regimes had identical slopes), Group-Buying outperforms posted prices for
intersecting demand curves, wherein the two demand regimes have different slopes.
Group-Buying does better under greater demand heterogeneity because it enables the
seller to set (non- linear) price-quantity schedules that optimize revenues under each
demand regime, thus maximizing his total expected revenues. With greater
heterogeneity in demand regimes, it is easier to induce a self- selective second-
degree
17
price-discrimination among customers, via quantity-discount schemes.

However, when the seller relies on posted prices, the single quoted price is
ineffective in exploiting the demand heterogeneity. Here, the seller is forced to make
a trade-off between revenues in one demand regime and revenues in the other.

17
See Tirole (1989).
16
Making this tradeoff leads to lower overall expected revenues as well. The posted
price mechanism makes this tradeoff in the case of parallel demand curves as well,
but since Group-Buying is also ineffective in enforcing price-discrimination (due to
the homogeneity of the demand regimes), posted prices perform as well as Group-
Buying in the earlier case.

We expect, therefore, that as the heterogeneity of the demand regimes increases,
Group-Buying would become more attractive relative to posted pricing, while
demand homogeneity would neutralize the advantage of pricing flexibility (enabling
second-degree price discrimination) that Group-Buying pricing mechanisms offer.
Since the heterogeneity in the two demand regimes stems from the difference in the
slopes of the two demand curves, and
1
2
m
m
m
· measures this difference, we expect
that the gains from Group-Buying increase in m. Proposition 3 below establishes that
this is indeed the case.

Proposition 3: The gains from using the Group-Buying mechanism increase as the
demand heterogeneity (m) increases.


Remarks:

1. Note that the demand curve given by q a mp · − is anchored at the point
( , 0) a on the quantity-axis, for all m. As the relative-slope parameter m increases,
this curve rotates downwards counter-clockwise (see Figure 4). Hence the
quantity-price coordinate at which the two demand curves intersect, given by
( , ) ,
1 1
a bm a b
q p
m m
− −
¸ _
·

− −
¸ ,
, moves down and to the right. As this happens, the
dominance of the demand curve q b p · − over the other ( q a mp · − ) increases;
this asymmetry of dominance translates of course to greater demand
heterogeneity. Thus, as the asymmetry of dominance between the two demand
regimes increases (i.e., demand becomes more heterogeneous), Group-Buying
schemes become more attractive relative to posted pricing.

5. Pricing and Production: Timing and Scale Economies

Our previous models implicitly assumed a certain sequence of events. For example,
the assumption of zero marginal costs is reasonable when production costs are sunk;
i.e., the commitment to production quantities precedes the pricing decision. When
pricing follows production, there are fewer degrees of freedom in the pricing
decision, under both Group-buying and posted prices. Perhaps the seller is unable to
exploit the richer space of price schedules under Group-Buying (compared to posted
prices), because of the limited flexibility in pricing. Put differently: under a different
sequence of events, with greater freedom in the timing of the pricing decision, Group-
Buying might outperform simple pricing. A related issue is the absence of scale
17
economies in our previous models. After all, since Group-Buying is an extension of
traditional quantity discounting, its greatest benefit might be precisely to maximally
exploit scale economies.

This Section throws the spotlight on these two issues: In the relative performance of
Group-Buying versus posted prices, (i) what is the impact of economies (or
diseconomies) of scale, and (ii) does the order of production versus pricing matter?
We analyze two different scenarios in this Section.

Scenario 1: Pricing follows production / procurement: This corresponds to the case
where the firm has already committed to production or procurement, and now needs
to determine the optimal price (in the case of posted prices) or price-quantity schedule
(in the case of Group-Buying). Thus, when the firm determines the price or price-
schedule, it has no control over production quantities. It may choose to sell less than
the entire quantity available. This scenario is realistic particularly under long
production or procurement lead-times, when the firm has to commit ex ante to
purchase or production quantities.

Scenario 2: Pricing precedes production / procurement: This corresponds to the
case where the firm has the option of deciding on the production quantity after
determining the quantity demanded for its price or price-schedule. This is feasible
when production or procurement lead-times are sufficiently small.

In a model such as ours, where the uncertainty is on the demand side rather than in
production, the firm will do better when pricing precedes production/procurement (which
is a form of production postponement), than when it has to commit early to production
quantities, under either market mechanism. Our interest, however, is in comparing the
relative performance of the two market mechanisms (posted prices and Group-Buying)
under each case, to see if the order of production and pricing affects the choice of the
market mechanism employed. We first analyze scenario 1, under the additional
assumption of constant marginal costs.

Case 1a: Pricing follows production / procurement under constant marginal
production costs:

We assume that there are two types of buyers (customers) in the market, and
N customers of each type. As N is a multiplicative factor that characterizes all revenue
expressions under both mechanisms and since we are interested in a comparative ranking
of the two mechanisms we will omit N from further analysis. It is evident from
inspection that this does not alter our results in any fashion. Each type of customer is
homogeneous in valuation, and buys at most one unit of the good. The valuation of each
type is drawn uniformly from the unit interval [0,1]; this is common knowledge. Thus,
one customer type has a higher value than the other with probability 1. We assume that
the supplier is a monopolist. Production is ex ante; i.e., the supplier commits to the total
production quantity before the revelation of demand through the price (or price-
schedule). We also assume that both the supplier and the buyers are risk-neutral, and
18
maximize their expected profits (value). We will first analyze the case of constant
marginal production cost: we normalize the marginal cost to 0.

We model the discovery of the optimal Group-Buy pricing schedule as a game between
the monopolist and the buyers. The sequence of events is as follows. The monopolist
determines his production quantity. Then he picks a price or price-schedule. Buyers then
choose to buy or abstain. The monopolist ‘folds’ the expected resulting demand into the
stage game(s) to arrive at the optimal price-quantity schedule. Recall that, under Group-
Buying, each customer’s bid creates positive externalities for all other customers (by
lowering the expected price). Hence, each customer needs to factor in the probability of
bidding by other customers (which is a function of their values), in determining her own
bid. A customer will buy (bid) if her value is greater than the expected price. Thus, a
customer may bid even when the current price under Group-Buying is greater than her
value for the good-- this may induce other customers to bid, further lowering prices. The
most general price schedule under Group-Buying for our problem may be represented as
[(p
1
,q
1
), p
2
], which translates to a unit price of p
1
if the quantity demanded is ≤ q
1
, and a
unit price of p
2
otherwise— the quantity discount is thus (p
1
- p
2
). The following Theorem
derives the unique (subgame-perfect) equilibrium for this game, under both posted prices
and Group-Buying. To solve the Group-Buying problem, we employ the Revelation
Principle (cf. Myerson (1991)). In the case of posted prices, a customer will buy the good
if her value is greater than the price.

Proposition 4: When pricing follows production, the unique subgame-perfect equilibrium
and supplier profits under posted pricing and under Group-Buying are as given below.
(i) Posted Pricing:
* *
0.5 and 0.5
p p
P π · ·
(ii) Group-Buying:
* * *
1 2
0.5 and 0.5
g g g
P P π · · ·
Thus the equilibrium solution as well as the supplier’s profit are identical under Group-
Buying and posted pricing.

Thus, the best that the supplier can do under Group-Buying with constant marginal costs
is to mimic the optimal posted pricing scheme— offering quantity discounts does not
improve the supplier’s profits. However, an argument can be made that our assumption of
constant marginal costs is suspect: as the academic literature previously surveyed (cf.
Monahan (1984), Lee et al. (1986), Kohli et al. (1989)) indicates, quantity discounting
(and, by extension, Group-Buying) is particularly effective under scale economies of
production or procurement. We now extend the previous analysis to the case of
production scale economies for the supplier.

Case 1b: Pricing follows production / procurement under production scale
economies:

The model analyzed here is identical to that of the previous case (with identical game
structure), with only one difference: the supplier’s production costs are given by c(q).
Since production precedes pricing, it is possible for the supplier to offer 0,1 or 2 units
under each market mechanism, depending on which of these is optimal. The production
19
quantity is a decision variable, but once decided upon, production costs are sunk for the
pricing problem. We do not impose any restrictions on c(q), although we would expect
that c(0) = 0, and c(.) is increasing in q.
18
The table below summarizes the results of the
analysis under both pricing mechanisms.

Market
Mechanism
Equilibrium
Solution for q=1
Supplier Profits for
q=1
Equilibrium
Solution for q=2
Supplier Profits
for q=2
Posted
Prices
*
1
3
p
P ·
( )
*
2
1
3 3
p
c π · −
*
0.5
p
P · ( )
*
0.5 2
p
c π · −
Group-
Buying
*
1
1
3
g
P ·
( )
*
2
1
3 3
g
c π · −
* *
1 2
0.5
g g
P P · · ( )
*
0.5 2
g
c π · −

Table 1: Solution and Profits for Case 1b

As Table 1 shows, for each possible production quantity (q = 1 or 2), the
equilibrium solution and supplier profits are identical under both posted prices and
Group-Buying. While the cost structure determines the optimal production quantity (0, 1
or 2) and profits under each mechanism, Group-Buying does not do better than posted
prices under every arbitrary cost-structure. In fact, the maximum profits under Group-
Buying are achieved by mimicking the posted price mechanism thereby resulting in a
single price equilibrium even under the Group-Buying scheme.

Thus, surprisingly, although Group-Buying is an extension of traditional quantity
discounts with coordination among multiple buyers, its advantage (if any) does not derive
from production (or, procurement) scale economies. However, one could argue that since
production costs are ‘sunk’ before making the pricing decision, the preceding model does
not provide adequate scope for the firm to exploit the generality of the Group-Buying
mechanism. More specifically, if production decisions could be made contingent on the
information revealed by the pricing mechanism, would Group-Buying outperform simple
posted prices? This question is addressed below by reversing the sequence of production
and pricing. We initially focus on the case of constant marginal costs, and then analyze
the case of scale economies.

Case 2a: Pricing precedes production / procurement under constant marginal
production costs:

This model is almost identical to that of Case 1a; the primary difference is that
the sequence of events is slightly altered. As before we model the discovery of the
optimal Group-Buy pricing schedule as a game between the monopolist and the buyers.
The altered sequence of events is as follows. The supplier first quotes a pricing schedule,
buyers respond with their purchase decision (demand is realized), and the supplier then
tailors his production to exactly satisfy his commitments (orders). In this case too the
optimal price schedule is discovered by backward- induction - the monopolist determines

18
Assuming that c(⋅) is increasing in q, a concave c(q) would correspond to scale economies, and a
convex c(q) would result in diseconomies of scale.
20
the optimal price-schedule that will lead to the optimal revenues in the terminal stage of
the game.

Such production postponement is feasible when production lead-times are short
enough and/or customers’ willingness to wait is long enough. Since the supplier now has
more options, the richness of the (non- linear) Group-Buying price schedule can perhaps
be better exploited, to extract and take advantage of demand information.

We also assume that the marginal production cost is c per unit, where 0<c<1.
As before we model the discovery of the optimal pricing schedule as a game between the
monopolist supplier and the buyers. The following Theorem derives the unique
(subgame-perfect) equilibrium for this game, under both posted prices and Group-
Buying.

Proposition 5: When pricing precedes production, the unique subgame-perfect
equilibrium and supplier profits under posted pricing and under Group-Buying are as
given below.
(iii) Posted Pricing:
*
1
2
p
c
P
+
· and
2
*
(1 )
2
p
c
π

· .
(iv) Group-Buying:
* *
1 2
1
2
g g
c
P P
+
· · and
2
*
(1 )
2
g
c
π

· .
The production quantities under each pricing scheme are determined by the realized
orders. Thus the equilibrium solution as well as the supplier’s profit are identical under
Group-Buying and posted pricing.

We see that, in this case also, Group-Buying does not provide any exploitable
informational advantage over simple posted pricing. Certainly, the supplier in our setting
does not garner higher profits from the ability to tailor production to the demand revealed
by the complex (non- linear) price schedule of Group-Buying. Finally we analyze the
more complex case of scale economies when pricing precedes production in the sequence
of events discussed below.

Case 2b: Pricing precedes production / procurement under production scale
economies:

We now analyze the case where (i) there are production scale economies and (ii) the
sequence of events is such that pricing precedes production. We assume that the cost of
producing i units is , for 1,2
i
c i · ; without loss of generality, we set
0
0 c · , i.e., there are
no production- independent fixed costs. Thus, production scale economies are induced by
declining marginal production costs. In our setting, the marginal cost of producing the
second unit (given by
2 1
c c − ) is not greater than the marginal cost of the first unit (given
by
1 0 1
c c c − · ), i.e.,
2 1 1
c c c − ≤ . The actual production quantities under each pricing
scheme are determined by the realized orders.

21
The following Proposition derives the unique (subgame-perfect) equilibrium for this
game, under both posted prices and Group-Buying.


Proposition 6: Under production scale economies, when pricing precedes production,
the unique subgame-perfect equilibrium and supplier profits under posted pricing and
under Group-Buying are as given below.

(i) Posted Pricing:
* 1 2
1 2
1
2 2
c c
p
c c
− +
·
− +
and
2
* 1
1 2
(1 )
2 2 2
p
c
c c
π

·
− +
.
(ii) Group-Buying:
* 1
1
1
2
c
p
+
· and
( )
( ) ( )
1 2 *
2 2 1 1 2 2
1 2
6 1 1
4 1 3 2 4
6 2 2
c c
p c c c c c
c c
− + ¸ _
· + − − + − + −

− +
¸ ,
and
( ) ( ) ( )
( )
3/ 2
2 2 2 3
1 1 2 1 2 1 2 2 2 1 1 2 2
1
52 72 36 15 18 9 12 2 2 1 3 2 4
108
GB
c c c c c c c c c c c c c π · − + − − + + − + + − + −

We find that in this (final) case, the profits for the monopolist from Group-Buying
dominate those under posted-pricing. As Figure 5 illustrates, Group-Buying does provide
an exploitable informational advantage over simple posted pricing.

[INSERT FIGURE 5 HERE]

The difference in seller profits under Group-Buying and posted pricing when
production costs change is a function of two factors: (i) Absolute profitability, which
under either market mechanism is clearly a decreasing function of production costs- when
costs (
1
c or
2
c or both) increase, the level of profits under both Group-Buying and Posted
Pricing fall. A drop in absolute profitability would tend to drive the profit difference
down as well. (ii) Scale economies, which also play an important role, as is obvious by
comparing the results of our analysis of cases 2a and 2b (Propositions 5 and 6).
Introducing scale economies while holding every other feature of the model constant
resulted in Group-Buying dominating Posted Pricing. To illustrate, setting
2 1
2 c c · ⋅ in
Proposition 6 leads to
2 1 1
( ) c c c − · , i.e., the marginal production cost of the first and
second units cost are identical. Plugging these values into the profit expressions yields the
profit expressions of Proposition 5-- the two market mechanisms yield identical profits to
the seller in this limiting case alone. Further, the degree of economies achieved due to
scale (which, in the absence of fixed costs, is measured by the concavity of costs, i.e.,
how steeply the marginal costs of production fall) is increasing in
1
c and decreasing in
2
c . This is easily seen, because the cost of the first production unit is
1
c and the cost of
the second production unit is
2 1
c c − . Thus, ceteris paribus, as
1
c increases, the first unit
costs more to produce and the second costs less—the relative production cost of the
second unit versus the first also falls. Similarly, as
2
c increases, the second unit costs
22
more to produce while the first unit costs the same, and the production cost of the second
unit versus the first rises.

From the preceding discussion, it becomes clear that when
2
c increases, both of the
above drivers- absolute profitability and scale economies- work in the same direction and
drive down the profit difference, as seen in Figure 5. A more interesting case is when
1
c
increases: also from the preceding discussion, the absolute profitability driver will cause
the profit difference
g p
π π − to fall, while the scale economies driver will cause the profit
difference
g p
π π − to increase in
1
c . Overall, the direction of change of the profit
difference when
1
c increases depends on which of the two drivers dominates. Figure 5
shows that, in fact, the difference in monopoly profits between Group-Buying and posted
pricing is an increasing function of
1
c for the entire feasible range of
1
c (and
2
c ). Thus,
surprisingly, the impact of scale economies driver dominates the absolute profitability
driver for the entire feasible range of
1
c , illustrating the significance of production scale
economies when pricing precedes production.
[INSERT FIGURES 6 & 7 HERE]

Figures 6 and 7 shed further light on the behavior of prices and profits under scale
economies under the two mechanisms. The cost parameter for the first unit,
1
c , is set to
one of three values: Low (
1
0.1 c · ), Moderate (
1
0.25 c · ) and High (
1
0.5 c · ). For each
fixed
1
c , the cumulative production cost for two units,
2
c , is varied along its entire
feasible range. Figure 6 plots the optimal prices under Group-Buying and Posted Prices.
It turns out that
*
1 2
P P P ≥ ≥ in all cases, with the equality holding exactly when
2 1
2 c c · ⋅ . In fact, P
1
is independent of
2
c , but an increasing function of
1
c . Intuitively,
the seller cares about the value of
1
P only when exactly one unit is sold, since this is the
only scenario under which
1
P affects his revenues. But then, the only cost that matters is
1
c . Both
*
1
P

and
*
2
P

are increasing in
2
c , and eventually converge to
*
1
P when
2 1
2 c c · ⋅ .
Under posted pricing, the seller has to balance out the benefit of sales at a higher price
(in the event of high customer valuations) versus the risk of losing sales at the higher
price (in the event of low customer valuations); this tradeoff is reflected in the optimal
posted price
*
P . Group-Buying affords the monopolist greater flexibility. Since the
probability of one customer having a high valuation and the other having a low valuation
is high, the seller sets
*
1
P to be high to trap the high value customer’s demand when one
customer has a high valuation and the other has a very low valuation. This risk is offset
by the safety net of a lower
*
2
P , to make the sales when both customers have moderate
valuations. Of course, Group-Buying has its drawbacks-- when both customers have high
valuations, the price settles down to
*
2
P , leaving money on the table, and when both
customers have low valuations, no sales are made. But these events have a relatively low
probability. Furthermore, the cost-savings from avoiding production pre-commitment
means that these losses are less than they would have been if production commitments
had to be made before orders are revealed. Figure 7 shows that for any fixed level of
1
c ,
23
the profit difference is greatest for
2 1
c c ≈ (maximum scale economies); it falls as
2
c
increases, and eventually becomes zero when
2 1
2 c c · ⋅ .

To summarize, when production can be tailored to meet revealed demand, scale
economies allow the exploitation of non- linear pricing under Group-Buying. Of the four
cases analyzed above, this is the only one in which Group-Buying outperforms posted
pricing. Clearly, revelation of demand information (through a non- linear price schedule)
and the optimal exploitation of that information (via sequencing production after pricing)
together play a role in driving our results. In the next Section, we throw the spotlight on
the role of product information as it affects the buyer directly (and the seller indirectly).
Information can induce customer heterogeneity, and may affect the relative performance
of the two price mechanisms.

6. Information and Valuation Revision

In this final Section, we model a different kind of uncertainty where buyers revise
their valuation of a product based on information received about the quality and attributes
of the product. When products are characterized by complexity, with variance in the
sellers’ reputation for service and reliability, buyers may rely on the information provided
by other buyers and product reviews available in online communities [Solomon (1999)].
Online reviews of products consist of both positive and negative information about the
product. Buyers process this information resulting in a change in their valuation of the
product. Some revise their valuation upward while others revise their valuation
downward. We model an additional layer of complexity in that the magnitude of the
impact of the information – on buyer willingness to pay is not known to the seller. There
are two possible states of impact, which we call high and low impact scenarios
respectively. Under the high impact scenario, the extent to which information affects
buyer valuations (in either direction) is given by the parameter θ. Under the low impact
scenario, the extent of impact is given by λ. For expositional ease, we model both
scenarios as being equally likely.

We assume there are n buyers all of whom enter the market with identical initial
valuations of a product given by a valuation V . As before they demand a single unit of
an atomic good. The sellers of the product provide information about product features
and quality ratings. Information about the reliability of the seller (reputation for prompt
service, accurate order fulfillment etc.) and the experience of other buyers with the
product is available at the sites of third party infomediaries (such as Epinions.com and
Amazon). Once buyers process this information, they revise their priors (valuations)
either upwards or downwards. The impact of the information in either (high or low
impact) scenario results in a fraction f of buyers revising their valuation upwards while
the remaining buyers revise it downwards. We call this ratio f the valuation revision
ratio. Under the high impact scenario, such upward revision of priors (by a fraction f of
all buyers) results in a valuation of V θ + while downward revision results in a valuation
V-θ. Without loss of generality, we normalize the values {V-θ, V θ + } to {0,1}
respectively. Under the low impact scenario, buyers are less affected by the information;
24
as before, they revise their valuations to V + λ and V- λ, where λ ≤ θ. When normalized,
this results in values:
¹
;
¹
¹
'
¹
+ − x x
2
1
,
2
1
, where
1
0
2
x ≤ ≤ . Since x is a measure of the
ratio
λ
θ
of the two magnitudes of impact (under the two scenarios), and λ ≤ θ, a higher
value of x corresponds to greater symmetry of impact between the high and low impact
scenarios ( ) λ θ ; , while a lower value corresponds to greater asymmetry of impact (i.e.,
λ θ = ) in the two scenarios. Hence we refer to x as the impact symmetry ratio. Note
that higher values of the impact symmetry ratio correspond to lower variance in
information impact, and vice-versa.

The following two Lemmas derive the optimal prices, and corresponding seller revenues,
under posted prices and Group-Buying.

Lemma 4: In a posted price market, the following are the optimal prices and the
resulting seller revenues:

1. )
2
1
)( 1 (
2
and
2
1
* then
f 6 2
f - 1

*
x f
n
x P x If − + ·
,
_

¸
¸
− ·
+
≤ π

2.

,
_

¸
¸
+ ·
,
_

¸
¸
+ ·
+
≥ x fn x P x If
2
1
and
2
1
* then
f 6 2
f - 1

*
π .


Lemma 5: Under the Group-Buy scheme, the following are the optimal prices and the
resulting seller revenues:
Case 1:
2f) 2(1
1

+
≤ x
*
1
*
2
1, if quantity demanded ;
Price *
1
- , otherwise;
2
q f n
p
P
x
p
≤ ⋅
¹
¸ _
¹
· ·
'

¹ ¸ ,
¹

and Revenue,
( )
* * *
1 2
1
, .
2 2
n
p p f x π
¸ _
· + −

¸ ,

Case 2:
1

2(1 2f)
x >
+


( )
* *
1 2
1 1
Price * { , },and Revenue, .
2 2
P x q q P f n x π
¸ _ ¸ _
· + ∀ · ⋅ +

¸ , ¸ ,


Finally, we compare the seller’s revenues from each of these mechanisms in the
following Proposition.


25
Proposition 7:

(i) The revenues to the seller from Group-Buying dominate the revenues from
posted price market when
( )
1
2 1 2
x
f

+
and
(ii) Revenues from the two mechanisms are equal when
( )
1
2 1 2
x
f
>
+
. In this
case, Group-Buying cannot do better than mimicking the posted price
mechanism.

Thus, Proposition 7 shows that the relationship between the two parameters—the
valuation revision ratio f and the impact symmetry ratio x— determines whether Group-
Buying does better than posted pricing or not. Further, this result is independent of other
parameters of the model. Figure 8 illustrates this result graphically.

[INSERT FIGURE 8 HERE]

Figure 8 shows that Group-Buying dominates posted pricing in markets
characterized by a lower buyer willingness-to-pay (where the valuation revision ratio f is
relatively low
19
). As buyers' willingness to pay increases, Group-Buying schemes are no
more attractive than fixed pricing schemes. Furthermore, as the impact symmetry ratio x
increases (i.e., variance of information impact falls), Group-Buying ceases to be the sole
dominant mechanism even under low values of the valuation revision ratio. Thus,
Group-Buying emerges as the dominant option only when the product is characterized by
(i) low willingness-to-pay on the part of the buyers, and (ii) high variance of information
impact.

This leads us to the important question: Why did the high-profile electronic Group-
Buying sites aimed at consumer aggregation fail without exception, over the past few
months? Our results shed some light on the reasons.

The items offered on sale on the B2C sites discussed in our introductory remarks as
well as in the Trade journals cited were mostly consumer durable goods; these are
characterized by medium to high information complexity [Notess (2000)]. For such
goods, buyers operating under ‘information overload’ (i.e., overwhelmed by the
combination of information complexity and bounded rationality ) often use only a subset
of the information available to them
20
[Lissack (1997)], or the ratings of other buyers and
online communities, to make the purchase decision [Appelman (2001)]. This leads to the
well-known herding effect, and, in effect, to high impact-symmetry (low variation in
information impact). From our previous analysis, we know that under high impact-
symmetry, Group-Buying would dominate simple posted pricing only for products for

19
Recall that if f is low only a small fraction of buyers revise their priors upwards while a large
fraction (given by 1 f − revise it downwards.
20
The reader is referred to [Simon (1979) , Simon(1991)] for further reading in this context.
26
which buyers have a low willingness-to-pay. Thus, Group-Buy sites found themselves
squeezed into categories of low- margin, relatively unprofitable products: this business
model was unsustainable. In fact, our survey of the popular Group-Buy sites revealed that
low-end (and low margin) consumer durables such as telephones, CD players, sunglasses
and Cassette Decks were the most popular items sold. In contrast, high end durables such
as luxury coffeemakers, Bose Lifestyle Home Theatre systems, high-end Nikon SLR
cameras and top-of-the-line Canon camcorders were conspicuously missing on the major
Group-Buy sites, while readily available through a variety of fixed-price sites.

7. Conclusion and Extensions

We began by investigating some of the reasons for operating Group-Buy pricing
schemes. We explored how sellers may use Group-Buy mechanisms to respond to
uncertainty in demand. Our model showed that under conditions of demand uncertainty
that commonly obtain, Group-Buy schemes often do not provide higher revenues to the
seller. In fact, when the seller manages to induce differential pricing (i.e., multiple price
points) by using quantity-based discounts, the Group-Buy scheme is often dominated by
simple pricing. In such cases, the best that the Group-Buy scheme can do is to mimic
posted prices, so that the seller’s revenues are the same under Group-Buying and posted
pricing.

While our analysis establishes that a monopolist cannot do better than a posted price
market with the Group-Buy scheme, in practice Group-Buy markets are likely to under-
perform posted price markets. In our analysis, we had not considered the erosion in
buyer’s valuation of the product from the inefficiencies surrounding the Group-Buying
process, resulting from its operational dynamics. Some of the inefficiencies associated
with the Group-buy scheme are as follows: A seller has to be certain about which demand
regime (and corresponding price) has been realized, before he can consummate the sales
to customers. For instance when a quantity
1
q q < is being demanded, he cannot be
certain whether the price will fall to
2
p or remain at
1
p . He has to wait a while before he
is able to ascertain the realized demand schedule and the price
21
. Group-Buying therefore
entails a waiting period for the buyers during which the seller has not determined the
equilibrium price. This delay and the resulting price-uncertainty often lead to utility
decay for the buyer, and since buyers can anticipate the delay (as was the case on
Mercata and MobShop) this results in an ex ante lower willingness-to-pay on her part and
thereby shifts the demand curve downward. In reality, due to coordination, waiting and
uncertainty related costs, buyers’ willingness to pay for these products tend to be lower
than in a posted price market where the transaction is instantaneously completed. As a
result, we expect to find that the Group-Buy mechanisms under-perform posted price
markets. The many examples from the trade and business presses of failures of Group-
Buying schemes confirm our findings.


21
For the seller to be sure that he can offer a price
2
p he has to wait until there is a build up in the
quantity demanded so that it exceeds
1
q (where
1 1 2 2 2 1
or :
l h
q a p q a p q q · − · − > ).
27
When buyers are heterogeneous, resulting in unusual and dissimilar demand regimes,
Group-Buying does offer some advantages over posted pricing. In the presence of both
demand uncertainty and heterogeneous demand regimes, the seller is able to use Group-
Buying to price-discriminate and capture some of the revenues lost by setting a single
price across both demand regimes. As the demand regimes become more and more
similar, the situation begins to resemble the cases analyzed in Proposition 1: the
advantage of Group-Buying over posted pricing shrinks and finally vanishes
22
.

We find that the value of Group-Buying as a price discovery mechanism depends on
the nature of the uncertainty about buyer valuations in the market. When the distribution
that characterizes buyer valuations is known beforehand, sellers are almost always better
off by running a posted price market—in fact, the only exception was when pricing (and
order revelation) preceded production and there were scale economies. However, when
the stochasticity with respect to demand is generated because buyers update their prior
valuations after consuming information about the product, Group-Buying strictly
outperforms posted prices for the seller, for a wide range of parameter values, and does as
well (by simply mimicking the fixed pricing scheme, and providing no volume discounts)
in the remaining range. Here the non- linear pricing schedule afforded by Group-Buying
enables precise exploitation of consumer heterogeneity through self-selectively induced
differential pricing. Unfortunately, the range wherein Group-Buy does strictly better than
posted prices (low values of both the Valuation Revision ratio and the Impact Symmetry
ratio) is also characterized by low buyer willingness-to-pay. As a result, Group-Buy sites
found themselves cornered into selling commodity- like low- margin products such as the
cheaper consumer durables.

Further, given the utility decay experienced by buyers due to the nature of the market
mechanism, it becomes even clearer that buyers that will be attracted to Group-Buying
site will be those that are price sensitive. This when combined with the fact that Group-
Buying is a dominant scheme only when a higher proportion of the buyers have lower
willingness to pay, results in a market whose revenue potential is considerably
diminished. Thus, sellers that operated Group-Buying sites were impacted by two
factors: the products that were sold on these sites were such that Group-Buying gave no
real advantage to the seller and when Group-Buying did emerge as the dominant market
mechanism, in such of these markets the seller was hit by adverse selection - i.e. the
conditions that made Group-Buying the dominant mechanism, made the market it itself
unattractive from the standpoint of profitability. Over the longer horizon, this business
model could not be sustained.

There are practical difficulties with Group-Buy schemes that we have not modeled
here explicitly but which can make the working of the scheme more inefficient. Firstly,
both buyers and suppliers face increased uncertainty about the final clearing price. This is
particularly a problem when customers are risk-averse, since their bids need to be based
on the expected price rather than the actual price. Secondly, the entire Group-Buying

22
For the advantage of Group-Buying to vanish, the demand regimes do not need to be identical. It is
enough if they are alike, as for instance if their slopes are the same, even though they may have
different offsets.
28
process typically takes from 10 days to a month. Typically, the sales and clearing prices
are finalized on a specific day committed to by the supplier, or when transactions cease
and the price stabilizes. Thus, customers have to incur the costs of delay in closing the
transaction. Thirdly, customers need to make, explicitly or implicitly, fairly complex
calculations in order to optimize their bidding thresholds, and to monitor the progress of
the Group-Buying constantly, in order to make their bids at the right time and price. (This
was reflected in our solution methodology, which finessed this problem by employing the
revelation principle.) Customers might experience a disutility from this buying process;
more importantly, their bounded rationality might lead to sub-optimal behavior-- and
hence, sub-optimal profits to the supplier. Future research should study the effect of these
factors on the effectiveness of complex market mechanisms such as Group-Buying vis-à-
vis simple posted prices.

Another possible extension is the analysis of Group-Buy mechanisms when there
is uncertainty about the quality of the product. When buyers can get signals about product
quality from the bids made by other buyers (similar to common values auctions), Group-
Buying mechanisms can be a powerful way for the seller to induce buyers to signal
product quality information to each other.

There are other possible approaches to effectively employing web-based market
mechanisms. It is important to point out that even if Group-Buying is not efficient as a
stand-alone approach, it might be an effective auxiliary scheme when used in tandem
with other market mechanisms such as posted prices and auctions. For example, the
market mechanism itself could be an effective price discrimination tool. The industry,
perhaps deterred by the initial wave of setbacks, has not so far experimented with such
combinations. Information has become increasingly recognized as a valuable commodity
and a strategic tool, particularly for web-based selling. Hence it is important for
manufacturers and distributors to control the dissemination of product-based information.
Infomediaries — online companies like CNet and PDA Buzz that trade in information –-
attempt to establish the primacy of the information they provide, in part by
commoditizing the physical products. Manufacturers and distributors can employ Group-
Buying and other complex market-mechanisms as a strategic response to reduce
competition and counter the commoditization of products induced by simple posted
prices in competitive markets.



29


References:


1. Appelman, H., "I Scream You Scream: Consumers Vent over the Net", The New
York Times, March 4, 2001.
2. Dennehey, M., " MobShop Closes Consumer Division"
http://www.auctionwatch.com/awdaily/dailynews/january01/1-011501.html
3. Kohli, R. and Heungsoo P., “A Cooperative Game Theory Model Of Quantity
Discounts”, Management Science, Vol. 35, No: 6, Jun 1989; pp 693-707.
4. Kreps, David, A course in Microeconomic Theory, Princeton University Press,
1990.
5. Lee, Hau L. and Rosenblatt, M., “A Generalized Quantity Discount Pricing
Model to Increase Supplier's Profits”, Management Science, Vol. 32, No: 9, Sep
1986, pp 1177-1185.
6. Lissack, M., "Of Chaos And Complexity: Managerial Insights From A New
Science", Management Decision, 35 (1997), 205 - 218.
7. Marschak, J. “Elements for a Theory of Teams”, Management Science 1,2, 1955,
pp 127-137.
8. Marschak, J. and Radner, R., Economic Theory of Teams, Yale University Press,
1972.
9. Mas-Colell, A., M.D.Whinston and J.R.Green, Microeconomic Theory, Oxford
University Press, 1995.
10. Monahan, J.P., “A Quantity Discount pricing model to increase Vendor profits”,
Management Science, Vol. 30, No:6, June 1984, pp. 720-726.
11. Myerson, R.B., Game Theory: Analysis of Conflict, Harvard University Press,
1991.
12. Nahmias, Steve, Production and Operations Analysis. Third Edition. Irwin Series
in Production Operations Management. McGraw-Hill, 1997.
13. Notess, G.R., "Web Wanderings: Consumers' Revenge: Online Product Reviews
and Ratings", EContent, April 2000.
14. Sandoval, G., Kawamoto, D., "Mercata Shuts Doors"
http://news.cnet.com/news/0-1007-200-4372403.html
15. Shankland, S., "Sun Micosystems Tries Group-buying Strategy"
http://news.cnet.com/news/0-1003-200-1851571.html
16. Simon, H.A., " Rational Decision Making in Business Organizations", The
American Economic Review, Vol. 69, No. 4. (Sep., 1979), pp. 493-513.
17. Simon, H.A., "The Journal of Economic Perspectives, Vol. 5, No. 2. (Spring,
1991), pp. 25-44.
18. Solomon., K, "Customer Reviews: It's All a Matter of Opinion", The Industry
Standard, October 11, 1999.
19. Tirole, Jean, The Theory of Industrial Organization. The MIT Press, 1989.
20. Value America Story: http://news.cnet.com/news/0-1007-201-1518733-0.html


30

Group-Buying Markets Cited in Section 3: Group-Buying: The Current State of Praxis


APPA: The Maryland Public Service Commission: http://www.md-electric-info.com/index.html
BBYT: BuildByte Construction Portal [Indian]: http://www.buildbyte.com/
COL: Chennai Online Bazaar (India): http://chennaionline.bazaare.com/
ECON: e.Conomy – operated by PriceWaterhouse Coopers: www.e.conomy.com
GBP: The Group Buying Partnership - http://www.groupbuying.co.uk/products.htm
IRPG: Independent Restaurant Purchasing Group: http://www.independentrestaurants.com/
LBI: Let’s Buy It.com – http://www.letsbuyit.com/lbisite/index.jsp
MPLY: McNopoly.com: http://www.mcnopoly.com/
MPSC: Maryland Public Service Commission: http://www.md-electric-info.com/info-
center/aggregators.html
OLC: Online Choice: Online Choice: http://www.onlinechoice.com/home/default.asp
PINE: The Printing Industries of New Englnad: http://www.pine.org/MS/groupbuying.htm
TBG: The Buying Group - http://www.the-buying-group.com
THM: The Happy Many: The Happy Many: http://www.happymany.com/index_en.html
STBZ: The StockBuzz Market: www.stockbuz.com

Vide Appendix-I for a more detailed list of Group-Buying markets.

31
Appendix – I: Group-Buying Markets


Business-to-Consumer Markets in the US:

Online Choice: http://www.onlinechoice.com/home/default.asp

McNopoly.com: http://www.mcnopoly.com/

e.Conomy – PreiceWaterhouseCoopers: www.economy.com

The Happy Many: http://www.happymany.com/index_en.html.

The Buying Group: http://www.the-buying-group.com/

LetsBuyit.Com (UK, Germany, France): http://www.letsbuyit.com/\

Chennai Online Bazaar (India): http://chennaionline.bazaare.com/

Ciao.com – (A German Group-Buying directory):
http://www.ciao.com/kategorien/1,202643,218630,55164.html

From Egypt.Com [Egyptian Group-Buying Market]:
http://groupbuy.fromegypt.com/users/Default.asp

IP Circles: http://www.contractsxml.org/ecap2000/students/FINAL/f-marko-
new/page3.html

Clust.Com [French Consumer Market]: http://www.clust.com/


Business-to-Business Group-Buying Markets


Independent Restaurant Purchasing Group (US):
http://www.independentrestaurants.com/

MaxGroup (US): http://www.maxgroup.com/Services/b2b_fremont.htm

Group Buying Partnership (UK): http://www.groupbuying.co.uk/

StockBuzz.Com: (Thailand): www.stockbuz.com

HCIS Group Buying: http://www.hcis.org/groupbuying.htm (B2B Health care
services).

The Printing Industries of New Englnad: http://www.pine.org/MS/groupbuying.htm

Printing and Graphics Communications Association:
http://www.pgca.org/pages/groupbuy.htm
32

Printing and Imaging Association of Mid-America:
http://www.piamidam.org/groupbuy.php

BuildByte Construction Portal [Indian]: http://www.buildbyte.com/

ShopeMates [Group-Buying Enabler]: http://www.shopmates.com/

BazaarE.com - Group buying services for businesses (and consumers):
http://www.rekha.com/cgi-bin/search/index.cgi?ID=981843428


Group-Buying Practices in Non-Profit Organizations

1. The Maryland Public Service Commission: http://www.md-electric-info.com/index.html

2. APPA: American Public Power and Environmental Agency
http://www.appanet.org/about/why/aggregation/f_buyingpower.pdf

3. The Center for Non-Profits: http://www.njnonprofits.org/groupbuy.html

4. Master Source Corp. : http://www.mastersourcecorp.com/




33
Appendix 2: Figures and Diagrams

Figure 1: Two Demand Regimes




Figure 2: Higher Price Point From High Demand Regime And Lower Price Point From
Low Demand Regime.

a
H
a
L
q = a
H
- p
q = a
L
- p
a
H
a
L
High and Low Demand Regimes
P
r
i
c
e
Quantity
a
H
a
L
P
1
P
2
q'
2
q
1
q
2
q'
1
a
H
a
L
P
r
i
c
e
Quantity
34
Figure 3: Higher Price Point From High Demand Regime And Lower Price Point From
Low Demand Regime.


Figure 4: Intersecting Demand Curves: Neither Demand Regime Dominates The Other
Universally.
a
H
a
L
P
1
P
2
q'
2
q
1
q
2
P
r
i
c
e
Quantity
a
H
a
L
(0,b)
(0,a/m)
P
1
P
2
q'
2
q
1
q
2
q = b - p
q = a - mp
(b,0) (a,0)
P
r
i
c
e
Quantity
35
Figure 5: Profit comparisons when Pricing precedes production under Scale Economies:
Group Buying (finally) does better than posted pricing for the seller. Also observe that
the difference
GB PP
π π − is increasing in
1
C and decreasing in
2
C .

















0.4
0.6
0.8
0.4
0.5
0.6
0.7
0.8
0.5
0.7
C
1
C
2
0.025
0.05
0.075
0.1
Difference in monopoly Profits between Group-
Buying and posted pricing (π
GB
- π
PP
)
36
Figure 6: Price comparisons under varying costs: Figure 6 consists of three graphs that
plot prices for varying costs.
1
c takes one of three values depending on the cost regime -
Low (
1
0.1 c · ), Moderate (
1
0.25 c · ) and High (
1
0.5 c · ) while
2
c varies
23
between
1 1
and 2 c c . P
1
and P
2
are the prices under the optimal Group-Buying schedule, while P
*

is the optimal posted price.









23
This is the range of values for
2
c that is feasible under production scale economies.
(a) Low Costs (C
1
= 0.1)
0.12 0.16 0.2
0.53
0.55
0.51
C
2
P
r
i
c
e
s
P
1
P
2
P*
0.5 0.6 0.7 0.8
0.55
0.625
0.7
0.50
C
2
P
r
i
c
e
s
P
1
P
2
P*
(c) High Costs (C
1
= 0.5)
0.3 0.4 0.5
0.52
0.57
0.62
C
2
P
r
i
c
e
s
P
1
P
2
P*
(b) Moderate Costs (C
1
= 0.25)
37
Figure 7: Profit comparisons under varying costs:
1
c takes one of three values depending
on the cost regime - Low (
1
0.1 c · ), Moderate (
1
0.25 c · ) and High (
1
0.5 c · ), while
2
c
varies
24
between
1 1
and 2 c c .







24
This is the range of values for
2
c that is feasible under production scale economies.
(a) Low Costs (C
1
= 0.1)
0.12 0.14 0.16 0.18 0.2
0.405
0.415
0.425
c
2
0.25 0.35 0.4 0.45 0.5
0.29
0.31
0.32
0.30
0.30
(b) Moderate Costs (C
1
= 0.25)
c
2
P
r
o
f
i
t
s
(c) High Costs (C
1
= 0.5)
c
2
0.5 0.6 0.7 0.8
0.18
0.19
0.21
0.22
0.23
0.24
0.20
Posted Pricing
Group-Buying
P
r
o
f
i
t
s
P
r
o
f
i
t
s
38
Figure 8: The two markets compared in terms of the two ratios that result due to buyers’
revising their willingness to pay.
0.2 0.4 0.6 0.8 1
0.2
0.25
0.3
0.35
0.4
0.45
0.5
I
m
p
a
c
t

S
y
m
m
e
t
r
y

R
a
t
i
o
Valuation Revision Ratio
Group-Buying &
Posted Price Markets are Identical
Group-Buying
Dominates
Comparison of Seller's Revenues Under
The Two Market Mechanisms

1. Introduction
Group-Buying schemes have been in vogue for very many years, particularly in the context of selling on television via the popular Home Shopping Network. Web-based variants of Group-Buying have received a lot of attention recently as part of the wave of innovative online market-based mechanisms such as auctions, reverse auctions and Priceline’s ‘name- your-own-price’ scheme. Nevertheless, unlike in the case of auctions (which has a rich history of analytical research over more than forty years), GroupBuying sorely lacks a theoretical framework. This paper develops analytical models of Group-Buying, building on extant academic research in the disciplines of Operations Management, Economics and Information systems. As we demonstrate, two streams of research — the quantitydiscounts literature in Operations Management and the literature on price discovery under demand uncertainty in Information Economics — serve as building blocks to develop a theory of Group-Buying. We compare the performance of Group-Buying with that of the market mechanism most widely used in electronic markets — simple posted prices — and identify the conditions under which each dominates. The remainder of this paper is organized as follows. In the next Section, we lay out our research objectives. In Section 3, we provide a comprehensive review of the current state of praxis, by surveying a variety of Group-Buying practices, world-wide, in the B2B, B2C and non-profit sectors. In Section 4, we review the relevant prior research and comment on important results that are relevant to our research. In Sections 5, 6 and 7 we develop three different analytical models of demand uncertainty and compare the performance of the two market mechanisms (Group-Buying and posted pricing) in each case. In Section 8, we discuss the managerial implications of our research and highlight possible extensions.

Section 2: Research Objectives
Our objective is to study the phenomenon, widely seen in current practice, of web-based Group-Buying. This is a specific instance of the proliferation of complex price-discovery mechanisms enabled by the Internet and the World-Wide-Web. Electronic Group-Buying market mechanisms seek to aggregate a variety of disparate buyers via the web— remotely and asynchronously— essentially by providing them price-based incentives to participate in the market and buy the good. In what is a classic volume-discounting scheme, the market operator offers buyers lower prices for higher volumes procured collectively. The Group-Buying markets that we surveyed were all characterized by prices declining in sales volumes. Further, web-based Group-Buying was almost entirely asynchronous— users have to work against some deadlines imposed by the seller, but have considerable leeway to make their bids via the internet within the week(s) during which the object is on sale. We provide a detailed exposition of several kinds of electronic Group-Buying markets in Section 3.

1

We know that simple posted prices are widely used in consumer sales—in traditional approaches such as retail and catalog selling, for instance, but also in web-based selling. While the web facilitates complex selling mechanisms such as Group-Buying, a larger question is how effective are such complex schemes, relative to simple posted prices, and more importantly, under what conditions do they add the most value? Our study is focused on answering two questions, both of immediate concern to current practice: (i) Given that a firm uses the Group-Buying (volume discount ing) market mechanism to sell its product(s), what is its optimal Group-Buying (price-quantity) schedule, and what are its revenues under this schedule? (ii) How does this optimal Group-Buying schedule compare with the other widely used market mechanism, specifically, simple posted prices? Specifically, what market conditions and product characteristics would justify the use of Group-Buying over posted prices? The practical application of insights into these two questions is obvious. To the best of our knowledge, our paper is the first in academia to (i) identify and delineate the Group-Buying market mechanism, (ii) develop analytical models of this mechanism, (iii) study the optimal such schedule under a variety of parameterized market scenarios, and (iv) analyze its performance relative to the simpler mechanism of posted prices (list prices).

3. Group-Buying: The Current State of Praxis
We quote the mission statements of two market operators that we surveyed to illustrate that volume discounting and demand aggregation are at the core of Group-Buying.
“Co-buying is co-operative shopping for the 21st century. It's a really simple way of getting better value by bringing people together via the Internet. By bringing together as many members as possible, LetsBuyIt.com can negotiate lower prices with merchant partners (also referred to as suppliers) or manufacturers. The more people, the lower the prices.”

-

Letsbuyit.com

“The more we are, the bigger the negotiation power towards trusted suppliers. The immediate result? Prices that fall without having to negotiate on your own. The more we are ("many"), the happier we are ("happy") because everyone pays less and everyone benefits from a better service.” - The HappyMany In our study of currently operational 3 online markets, we found that Group-Buying (referred to as ‘co-buying’ in the above quote) is a widely deployed price-discovery mechanism in a variety of markets and contexts. These markets are not limited to a particular geographic region (such as the United States) or a particular product category. We will see that the Group-Buying mechanism is extensively used in the United States,
3

As of mid-September 2002.

2

the required number of partic ipant Co-buys is not reached). in both business-to-business (B2B) and business-to-consumer (B2C) markets. GBP] 4 and supplier [LBI. 4 Capital letters in square parentheses refer to websites of companies listed in ‘Group-Buying Markets: Citation List’ provided at the end of this document. telecommunications & connectivity.Europe (including in Germany. furniture. computer hardware. or that employ theoretically interesting variants of Group-Buying. Members are assured of an initial maximum price (a ‘price ceiling’). If you choose to buy at the Best Price and the Best Price is not reached (i. lowest price. Given the large number of active Group-Buying sites (well over 50). consumer electronics. TBG. interesting examples of this deployment.Buying websites from the B2B. France and the UK) and Asia (including in India. is a consumer groupbuying market currently functional in the UK. commercial print.e. and company travel. B2C and non-profit sectors is provided in the Appendix at the end of this document. While we documented markets in three continents that deploy the group-buying mechanism. incrementing the total volume demanded. Demand is aggregated across multiple buyers. that prominently tout the benefits of Group-Buying to their members. All buyers pay the same market-clearing price. that works in a manner very similar to the e. In the words of Let’sBuyit. Evidence of this belief is adduced by the web-sites of both buyer [IRPG. however. you will receive the product (provided your payment details are accepted) and you will pay the closing price. software. an exhaustive review is beyond the scope of this paper. The Group-Buying site e. This review discusses representative. THM] consortia in a variety of markets. Germany and France. Let’sBuyit. As they place orders. food and wine. APPA. If you choose to buy at the current price. A nuance to the standard Group-Buying mechanism that this company permits is that buyers can choose not to declare a price ceiling—buyers’ greater risk without a price ceiling may be offset by guaranteed availability of the good at the common. This B2B market allows buyers to purchase indirect goods and services such as office supplies.conomy.or at the Best Price only.conomy model. and jewelry. Singapore and Egypt). 3 .com. you need to decide whether you would like to buy at the current price – the price reached when the Co-buy closes . and prices decline with increasing volumes until they reach a stable level. operated by PricewaterhouseCoopers . we have restricted the discussion below to some interesting instances that are representative of several currently operational markets. is a successful market based in the US [ECON]. and in public as well as private markets. Group-Buying schedules are employed for branded consumer products as well as intangible services such as bandwidth and network security. of relevance to our modeling-based analysis. the resulting pricedeclines are broadcast (and available to all) until the market clears at a predetermined time. Thailand. ‘temps’ (temporary staffing services). sport and leisure. These markets are characterized by the belief that both suppliers and buyers stand to benefit through Group-buying. at substantially reduced prices that result from higher purchase volumes. “…Before you join a Co-buy. quoted in the beginning of this Section. a fairly comprehensive list of 26 representative Group.com. you will not receive the goods. Your order will be cancelled and you will not be charged…” Product categories sold in this market include exercise equipment.

Another version of the Group-Buying mechanism called StockBuzz (run out of Thailand) is operated by Asian producers of high quality yarn who sell their products to manufacturers of high-end apparel [STBZ]. This electronic group buying market (run out of India) is a private market that allows buyers to place orders by specifying price-quantity schedules where purchase is required only if the supplier is able to meet the price specified by the buyers. and passes these discounts on to buyers [GBP]. while production and inventory stocking decisions precede pricing in the first category (modeled in Case 1. Using the two estimated demand schedules as guide rails. retailer feedback. computer peripherals. consumer electronics and electrical appliances. past sales data. the buyers are obliged to follow through on their commitments. based on prior seasonal data and other available information. In this case. Then. Section 4 of our paper). Thus. the second category is an example of production postponement (the Case 2. In this case. most importantly. The mechanism works through a process of demand aggregation wherein buyers commit to purchase quantities subject to certain price ceilings-. Buyers (apparel manufacturers) are allowed to place conditional orders which they are then obliged to honor should the price reach the levels specified in the order contract. they commit to a schedule of prices for different quantities with prices declining in quantities. which tends to fluctuate between a robust (high demand) regime and a weaker regime reflective of an adverse consumer demand cycle. The apparel manufacturers forecast demand based on a number of macroeconomic factors. In a second category of goods that includes kitchen appliances. A different flavor of this practice can be found in the market called Chennai Online . Since most of the manufacturers sell to the same retailers and in the same urban markets. the sequence of demand solicitation and production (or procurement) of the goods is reversed. The yarn producers in turn face uncertainty in demand. 4 . which targets buyers who are redistributors [COL]. computer equipment. the supplier announces prices at different levels of demand (price-quantity schedules) and the buyers place bids serially until the market clears at a particular price. Thus the yarn producers estimate the probabilities of these two market states. magazines and periodicals from the popular press and computers (servers and PCs). Section 4 model). Section 2 of our paper). their demand estimates tend to be highly correlated. suppliers either produce the goods (kitchen appliances and computer peripherals) or buy them from third parties (consumer electronics and electrical appliances) in exact quantities. office supplies and vehicles. yarn suppliers use Group buying as a response to uncertainty about which of the two states of demand will be realized (modeled in Cases 1 & 2. They claim on their web-site that “Group Buying offers a simple and cost effective opportunity for small to medium sized businesses to combine their individual buying power for the overall benefit of each individual member” [GBP]. This is a direct correlate of purchasing commitment conditional on demand realization modeled in our paper. In the case of products such as software suites for office use.The Group-Buying Partnership in the United Kingdom is a consortium of small and medium sized businesses which negotiates volume discounts from suppliers of items such as electricity and gas. seasonality and. What is interesting in this context is the order of events that lead to price discovery. and reflect consumer demand cycles. suppliers commit to a group-buying schedule and demand is realized when a stable price-quantity tuple is reached.should the consortium succeed in procuring the quantities at or below these price ceilings.

If there is enough interest. We found that the Maryland Public Service Commission supports individual buyers in their efforts to form buying groups to negotiate better rates for higher aggregate levels of power consumption [MPSC]. the American Public Power Association and the Environmental Action Foundation have launched a number of Group-buying initiatives that bring together consumers to negotiate lower rates with power utilities across the US [APPA]. Smaller restaurants often find it difficult to meet these minimum criteria. IRPG argues that the benefits of its Group-Buying scheme. those consumers whose bids for the products were higher than the market-clearing price are required to buy the products.Group-buying mechanisms are also observed in specialized markets. Two companies. would probably not meet the manufacturer's minimum volumes. No matter how many restaurants you have the IRPG can add strength and stability to your purchasing power. In the hotel and restaurant industry the Independent Restaurant Purchasing Group (IRPG) brings together over 3000 restaurants across the United States and negotiates volume-based discounts with suppliers on their behalf [IRPG]. and shares the benefits of the lower prices with the buyers (here again. Firstly. the market operator refunds the buyers’ paid. As IRPG argues. Many firms operating in the non-profit sector are also using Group-Buying. for its members are two. the others exit the market. by themselves. A noteworthy feature seen in this market is the pre-order feature wherein buyers ‘express an interest’ in a product that is not currently featured in the market. when combined with those of the entire membership they exceed the minimum and in many instances qualify for further "multi. The requirement that buyers commit to a quantity and price by paying the amounts upfront ensures that buyers do not make frivolous bids.up amounts. However. buying power is enhanced for large chains as well as small restaurants: “Our national and regional contracts are based on the total volume of our group versus that of an individual restaurant. If trade is not possible. If the market operator succeeds.). Branded consumer goods are also sold via the Group-Buying mechanism. he charges all buyers (who are willing to pay the clearing price or more) the same floor price.fold. OLC].unit-allowances" normally only available to large chain accounts. In the skin-care product market. called the IRPG Volume Rebate Program. The market operator then negotiates a price from suppliers for the aggregate quantity. at prespecified prices. 5 . Similarly. the market operator (in this case. the market operator ‘The Buying Group’ [TBG] operates a relatively straightforward groupbuying scheme where buyers commit to quantities at a maximum price. If their suppliers offer discounts that support a market-clearing price. “Most restaurants. McNopoly and Online Choice. many manufacturers insist on minimum volumes for doing business with them.” Secondly. we see ‘procurement postponement’. out of efficiency considerations arising from their production and transaction scale economies.” [IRPG]. and also provides an incentive for the market operator to negotiate with suppliers. The Buying Group) attempts to procure the product at a price that will make trade possible for at least some buyers. operate markets that allow consumers to bid for specific branded items from a menu of offered items [MPLY. studied in Section 4 of the paper. and pay for their desired quantities upfront.

Group-Buying is very often targeted towards buyers with low bargaining power-. These are discounts you may not be able to receive if you were to negotiate directly with each company” [PINE]. and verifies the extent to which these underlying beliefs are true under various market conditions. In fact. explicitly and prominently claim that ‘enhanced buyer bargaining power’ is the single biggest advantage of Group-Buying. Ceteris paribus.Lesser the price. A key feature that sets all GroupBuying schemes apart from other market mechanisms is suppliers’ beliefs that precommitting to a price-quantity schedule where the prices are monotonically declining in total purchase quantities (and not just an individual buyer’s purchase quantities) will maximize supplier revenues by inducing greater buyer demand. PINE combines the overall buying power of its 513 member companies to negotiate these discounts. many websites across all types of products. “PINE has made arrangements with several companies — companies you use in running your business — to provide discounts on their products and services. However. in some cases. The Group-Buying Mechanism: Theoretical Underpinnings The Group-Buy business model has two components: (i) a quantity discount scheme offered by the seller. the former approach leads to higher supplier profits but is not always feasible due to long production/procurement leadtimes. The more members that join a Co-buy. BuildByte [BBYT] puts it even more succinctly.” [LBI]. services and countries. two streams of literature 6 . Our paper aims to provide a more analytical response to this discourse. More generally. who need to coordinate their actions to the extent possible. the lower the price becomes…. and then produce (or procure) the products in volumes exactly calibrated to the realized demand. the important practical question of whether to adopt Group-Buying when production postponement is or is not feasible hinges on the relative performance of Group-Buying vis-à-vis posted pricing in these two cases. Thus. In other cases. seemingly ‘commonsensical’ assumptions (reflected above in the quotes from various companies). The market operator Let’sBuyIt. suppliers first produce (or procure) products and then sell via a Group-Buy market through a volume discounting mechanism. Similarly. Thus. and (ii) decentralized decision making by the buyers (consumers). the widespread use of Group-Buying in practice is predicated on a number of unverified. Since the answer is by no means obvious. We found that.To summarize. “Bigger the volume .” We saw that the Group-Buying schemes also differ in the order of events that lead to price discovery.individual consumers or small to medium businesses. The construction industry’s portal site in India. through the artifice of mathematical modeling. our model and analysis in Section 5 attempts to shed light on this question. the Group-Buying intermediary Printing Industries of New England (PINE) asserts on its website. we find evidence of widespread deployment of Group-Buying as a pricediscovery mechanism in both B2B and B2C sectors. 3.com summarizes this belief: “We make it easy for our members to come together online to benefit from collective purchasing power. suppliers first commit to a price schedule leading to demand realization.

and study the praetor-efficient outcomes of different axiomatic bargaining solutions due to Nash. the discount applies only to additional units beyond the breakpoint (see Nahmias (1997) for examples of each). in larger batches.1 Quantity Discounts Quantity (or Volume) discounts have a long and well-studied history in the context of trades between businesses. (1989) make the strong claim in their paper that “from a transaction-efficiency perspective. The usual structure of such schemes is that there are breakpoints defining changes (decreases) in the unit cost. MobShop and other players. this finding (which applies to quantity discounts in a traditional context) does not hold for the case of Group-Buying. The more popular all. or the buyer-seller system. The supplier offers a quantity discount scheme that optimizes over two different effects: (i) the buyer’s actual orders (as in Monahan (1984)). who are presumably those with higher expected utility from the good. Lee et al. 7 . the discount is applied to all the units in a given order. This lowers prices and encourages other custome rs to follow suit. Kohli et al.are particularly pertinent to our analysis: (i) Quantity Discounts. Lee et al. under the incremental discount scheme.units discount scheme is relevant to Group-Buy schemes employed by the Home Shopping Network. (1989) analyze quantity discounts as the outcome of cooperation (through bargaining or negotiation) between buyer and supplier. or increases the optimal order quantity over that without quantity discounts. The guarantee that early bidders. and Eliashberg. The purpose of quantity discounts is to encourage the buyer(s) to purchase more of the seller’s good. Kohli et al. the seller. at least. Monahan (1984) extends this idea by deriving the optimal pricing schedule that a supplier should offer. In fact. This induces EOQ-like buyer behavior.” As we will demonstrate in the paper. (1986) extend the analysis to the case where the supplier can use a different lot size from the buyer. Kalai and Smordinsky. Thus. (1986). since both production batch-size and frequency are now decision variables. the choice between incremental and all-units discounts is a matter of firm or industry practice. These are individually discussed below. Kohli et al. 3. which determine his production and inventory costs. and (ii) the Coordination Problem. the EOQ trades off set-up costs and holding costs. and (ii) the supplier’s actual production schedule. where the outcomes for both buyer and seller critically hinge on the structure of the discounts offered. While Monahan (1984) assumes that the supplier produces in lot sizes that mimic the buyer’s orders. given that the buyer subsequently optimizes her profits by ordering appropriate quantities. There is a rich stream of academic literature in Operations on quantity discount schemes [cf. consumers face a coordination problem. Monahan (1984). not a result of their desirability for the buyer. The Economic-Order-Quantity (EOQ) formula gives the optimal order quantity for a buyer who has scale economies in procurement or production. promotes their entering the market early. Under all-units. There are two popular forms of quantity discounts—all-units and incremental [Nahmias (1997)]. This makes the supplier’s analysis more challenging. or. (1989)]. it is very important under the GroupBuying scheme that early bidders are not penalized by higher prices. will pay the lowest price offered to any buyer.

since the final clearing price 5 There is a vast literature on externalities. or traditional channels such as bricks-and. (1986). the players are said to be a team (Marschak (1955). Lee et al. Clearly.g. Economists have analyzed the use of nonlinear pricing such as quantity discounts for second-degree price discrimination [Tirole (1989)]. the customers under Group-Buy are not operating as a team. since his bid can only benefit other consumers by lowering the expected clearing price. Group-Buying involves both waiting and uncertainty on the part of the buyers before they know the outcome (the market clearing price). a discussion of which would take us too far afield from the focus of this paper. high quality items by low margin items. In contrast to such negative externalities. using web-based posted prices. The classic example of a negative externality is the presence of congestion effects that dampen system performance: each additional customer increases the system congestion and inflicts additional delay costs on all other consumers. each incremental customer inflicts an extra cost (or loss of benefit) on all the others. and is discussed in most graduate textbooks in economics (cf.mortar retailers. This reduces the cannibalization of the sales of high margin. 5 Under negative externalities. Mas-Colell et. they maximize their expected individual surpluses from their purchases. and more generally for all web-based models that discharge the function of demand aggregation. the final clearing price depends on both the number of buyers who arrive on the site and their willingness to pay for the product. Secondly. Marschak and Radner (1972)). al (1995)). (1989)] is their focus on quantity discount schemes as an instrument to improve transaction efficiency.2 The coordination problem As discussed previously. electronic items) that are less than state-of-the-art. However. Firstly. a distinguishing feature of the Group-Buy business model is that the traditional quantity discount model is overlaid with a coordination problem among consumers. Many of the Group-Buy sites offer products (e. 3. lower.valued (patient but price-sensitive) customers could be served through Group-Buying. While this may be part of the proffered rationale for Group-Buy schemes. The famous ‘public good problem’ of economics is an example of this. The pure coordination problem occurs when all players optimize a common objective function. the sale has to be kept open long enough to ensure that a sufficient number of buyers arrive at the site. each consumer’s behavior (whethe r bidding or not bidding. Kohli et al.A common feature of the academic literature discussed above [Monahan (1984). Available evidence from the business press on this subject support these hypotheses. since they are willing to incur the additional delay that Group-Buying entails in return for a lower expected price. and the timing of the bid) affects other consumers’ surplus. higher. Thus.valued (impatient) customers could be served through retail outlets or web-based posted prices. Even in the absence of such quality-based second-degree price discrimination. in this case. while the latest product is sold in more conventional ways. Group-Buy sites could simply serve heterogeneous customers through multiple channels. the literature on externalities is of relevance to the Group-Buy problem. 8 . Further. the coordination problem in the case of Group-Buy is more benign – a bidding consumer induces a positive externality on all other consumers. If there is a strong negative correlation between customer willingness-to-wait and their valuation of the good (as is often the case). at steep discounts. the major use in practice of Group-Buy channels has been for price discrimination.

causing the price to fall below his reservation value. when the nature of the demand (function) is known. Thus. almost perfectly aligned. The monopolist cannot observe or infer the demand regime prior to the pricing decision. he can do as well by offering a fixed price that maximizes his expected revenue. rather than the incremental discount model). The supplier can and should build this factor into his discounting schedule. All buyers get the good at the settlement price. As discussed earlier. Under certain conditions. Buyers arrive at the market and declare their intent to buy at the current price point. and simply announce a posted price equal to that settlement price. Clearly. This is because.is the same for all consumers (analogous to the all-units quantity discount model discussed previously.. the seller can calculate the exact settlement price from a Group-Buying scheme.1 Model of Parallel Demand Regimes We construct a theoretical model of a monopolist-seller facing demand uncertainty.like’. The monopolist seeks to sell a quantity Q . 4. The demand for the 6 The term ‘externalities’ here should not be confused with ‘network externalities’. Network externalities are just one of many different kinds of externalities. the consumer will bid for the good even above his reservation value. The monopolist operates in one of two demand regimes (which we term high / low) each of which is equally likely. this feature of the Group-Buy scheme distinguishes it from all other market mechanisms. where the price drops as quantity increases. As the quantity demanded exceeds a predetermined (and preannounced) level. in a Group-Buy scheme the seller declares a price-quantity schedule. The seller has an estimate of the demand for his product (at every possible price) but does not know it exactly. the consumers’ incentives are ‘almost team. We analyze the impact of different kinds of demand uncertainty on the seller’s pricing strategy and compare the performance of the two pricing mechanisms (Group-Buying and posted prices) under each. since his bid may induce others to bid. he will not resort to a Group-Buying scheme. the price drops to the next lower price point. i. 4. Web-based Group-Buying schemes are predicated on offering buyers discounts for higher volumes of purchases. Each buyer demands exactly one unit of an indivisible good. in expectation. This in turn encourages more buyers to bid. The appropriate schedule can trigger an avalanche of bids due to positive externalities 6 .e. which is the price point at which the price stabilizes. It is only when the seller is not sure about the structure of the demand that he would resort to Group-Buy schemes. the settlement price is not more than (and usually less than) buyers’ bids. which is a demand side phenomenon under which the installed base of users of a network experience gains in value when the size of the network expands due to adoption by new buyers. 9 . notably posted prices and auctions. Models of Demand Uncertainty A monopolist’s choice of pricing mechanism is often driven by the nature of the demand uncertainty that he faces. The market consists of a fixed number of buyers transacting within a single period. When the seller knows the structure of the demand with certainty. the consumer may be better off by bidding for the good above his reservation value.

With the above formulation we turn to the monopolist’s revenue maximization problem under Group-Buying. ( q2 . We assume that demand for the good is linear. Since there are only two demand regimes. p1 ) . the monopolist will pick a pair of price-quantity tuples [( q1 . the demands are given by: ' ' Qh ( p ) = ah − m ⋅ p and Ql ( p ) = a l' − m ⋅ p respectively. and given by: Q ( p ) = a −m ⋅ p. a general feature of all Group-Buy schemes is that the price is decreasing7 in the total quantity demanded. where the price schedule consists of two or more price points corresponding to different quantities demanded. We preserve this feature in our model. Further. p1 ) . It will be clear from inspection that the solution and the ordering of market mechanisms by revenues will not change if a positive marginal cost were to be factored into the analysis. such that the expected revenue is maximized. the availability of complementary or substitute products and a host of macroeconomic factors. 9 Under the high and low demand regimes. subject to the condition that q2 > q1 and p2 ≤ p1 . ( q2 . As discussed above. The two demand regimes are shown in Figure 1 below: the high demand regime is the outer demand curve. The notation [( q1 . or equivalently. our analysis assumes a non-zero marginal cost. the offered price 8 should be decreasing in the quantity ordered. 10 . the marginal cost might matter: in this case (studied in a later Section). we normalize m to 1 and the resulting demand curves are given by: q = ah − p and q = al − p. We distinguish between offered and realized prices here since a single price may be realized which is one of the many offered prices. 9 The linear demand curve is of course widely used in the Economics literature. pi ) and ( q j . Tirole (1989)). When pricing precedes production. and p2 when the total sales quantity is > q1 7 8 We do not assume monotonicity although monotonicity would strengthen our results. with one tuple corresponding to each regime. we are justified in treating the marginal cost as sunk and therefore maximizing revenues (as opposed to revenues net of cost of production). We will consider the case when the two regimes have non-parallel demand curves in models that follow. p2 ) ] .monopolist’s product is determined by a number of factors. including product attributes and consumer preferences. This formulation results in parallel demand curves for the two regimes. He has to choose from one of two pricing mechanisms: Group-Buying or posted prices. Without loss of generality. where ah > al' . 10 Since production precedes pricing in our model. p j ) : q j > qi ⇒ p j ≤ pi . to capture the relationship between prices and quantities (cf. we need that ∀ ( qi . p2 ) ] means that the seller charges the price p1 when the total sales quantity is ≤ q1 . [INSERT FIGURE 1 HERE] Thus the monopolist’s problem is to pick a price-quantity schedule that will maximize his expected total revenue 10 without knowing which of the two demand regimes is realized.

the price-quantity pair that is operationalized is the one that the monopolist intended for that regime. Consider the following two scenarios: If the lower demand regime is realized. If the higher demand regime is realized. under each regime. Thus. The implementation of this idea. prefer the price-quantity pair intended for them. 14 These constraints result from the definition of Group-Buy mechanisms: higher quantities lead to lower prices. in connection with the well-known revelation principle of economics. then seller would want the price to remain at p1 . to ensure that under each regime. The quantity q1 acts as a limit for the higher price p1 . al (1995) or Myerson (1991). the price that will prevail is p2 and the quantity sold will be given by q2 = a l − p 2 . This would imply that he expects to realize a higher quantity of sales from the lower demand regime 12 . The seller will offer the following pricing scheme 13 :  p1 if quantity q ≤ q1 = ah − p1 . 11 will become clearer in the exposition below. Under this scheme the price (outcome) will always be p2 irrespective of the demand regime that is realized. The seller has two alternatives open to him: (i) pick the higher price point from the high demand regime (and therefore the lower price point from the lower demand regime). 11 . [INSERT FIGURE 2 HERE] In Figure 2. 13 It will be clear by inspection that this pricing scheme maximizes the seller’s revenues under Case 1. q1 = ah − p1 and q2 = a l − p2 . see Mas-Colell et. Lemma 1: The seller’s total revenue in expectation when he chooses the higher price point from the higher demand regime is given by: 11 For an excellent ex position of the Revelation Principle.but ≤ q2 . (ii) pick the lower price point (and therefore higher sales quantity) from the higher demand regime and the higher price point (and therefore lower sales quantity) from the low demand regime. while maximizing their own value. As a result. Case 1: Higher price point is picked from the high demand regime (see figure 2). This leads us to our first result about the seller’s optimal revenue. Figures (2) and (3) illustrate options (i) and (ii) respectively. the equilibrium outcome should be that the customers. rather than the alternative. Further. however. 12 A higher price implies a lower quantity and vice versa in a Group-Buy pricing schedule. Price P =   p2 otherwise. incentive compatibility constraints have to be satisfied. buyers in this demand curve would demand a quantity given by q2 = ah − p2 which is greater than ah − p1 . the seller would have to offer the lower price corresponding to the higher quantity level. subject to the constraints14 p1 ≥ p2 and q1 ≤ q2 .

1) Remarks: 1. the revenue-maximizing prices in the high and low demand markets would be. Price P =   p2 otherwise. 15 Proof of this and all other propositions can be found in the additional document entitled “Proof Of All Results”. the price does not slide down to p2 . respectively. We do not specify what p1 should be since this price is never realized. a a a a * * * * p1 = l and p2 = h . and the optimal price P* is given by P* = ah + al 15 . 2. this solution is infeasible under the Group-Buy mechanism. By assumption of this Case. 12 . p1 ). if the seller could maximize revenues individually in the two markets. ( q2 . 2 2 2 2 However. with the corresponding sales quantities q1 = l and q2 = h . which ensures that. since the constraint that the price should be decreasing in quantity demanded is violated (since * * * * p1 > p2 and q1 > q2 ). in the event the low demand regime is realized. we need to maximize the seller’s revenue within the feasible set of prices. Since separate maximization of revenues across the two demand regimes leads to an infeasible price-quantity schedule. The above scheme is in essence a single price scheme (fixed price) since only a single price will prevail irrespective of which demand is realized. Case 2: The seller picks the lower price point from the high demand regime and the higher price point from the low demand regime. 4 (3. Now the separator between the two pricing points is the quantity q '2 . [INSERT FIGURE 3 HERE] In this case. subject to the constraints p1 ≥ p2 and q1 ≤ q2 . Lemma 2 provides the optimal solution. the seller will offer the following pricing scheme:  p1 if quantity q ≤ q1 = al − p2 . the only constraint on p1 is that p1 > p2 . In this case separation of prices across the two demand regimes seems possible.π ( q1. 16 Of course. 16 Recall that the seller’s objective is to implement a different price in each demand regime to optimize his total expected revenue. p2 )   * (a + a ) = h l 16 2 .

( ) Remarks: Thus the monopolist is unable to exploit the Group-Buying price schedule to implement discriminatory pricing based on which demand regime is realized. by setting his price p * = p* (= p2* ) . This implied that at any price the quantity demanded under one demand regime dominated that demanded under the other. ( p2 . We now consider demand uncertainty where one demand regime does not universally dominate the other.   8  2  In particular. the monopolist is able to offer a separation of prices based on quantity demanded. the seller’s revenues from Group-Buying can never exceed his revenues from simple posted prices.2 Model of Intersecting Demand Regimes In the previous section. q1 ) . the monopolist can clearly do as well using posted prices. Any deviation from posted pricing (such as differential pricing to discriminate between realized demand regimes) will lead to sub-optimal revenues for the seller. we study the case when the curves for the two demand regimes intersect: one regime may result in a higher quantity demanded for one range of prices. q2 )  = h . The seller’s revenues under Group-Buying and posted prices are 1 identical. Proposition 1: Under the above model of demand uncertainty. we considered the case of demand uncertainty wherein the two demand regimes were parallel. q1 ) . Specifically. Remarks: 1. the price is independent of the demand regime). the optimal (revenue-maximizing) Group-Buying scheme simply mimics the optimal posted price. but the revenues from posted prices dominate the resulting revenues from Group-Buying. However. when the two revenues are equal it is important to note that the Group-Buy scheme does in effect act as a posted price market. The following Proposition derives from Lemmas 1 and 2. if al ≥ ah 2 −1 . ( p2 . Since * * p1 = p2 under the optimal Group-Buy scheme (i. q2 )  = . otherwise. The above analysis shows that the revenue from Group-Buy schemes may at best equal the revenue to the monopolist from posted pricing.    4 16  2 a a  * * p1 = p2 = h and π G ( p1 .. 4. In Case 2. In Case 1.e.Lemma 2: The optimal (revenue-maximizing) prices and seller profits under groupbuying are given by  * al + ah ( al + ah ) 2 *  p1 = p2 = and π G ( p1. In fact. any feasible group-buy schedule that sets the prices such that p1∗ ≠ p2∗ yields revenues less than the above solution. thereby establishing a single-price equilibrium across the two demand regimes. while the 13 . the price p1 always slides down to p2 .

2) p  a  2   . Let m = 1 .other regime may dominate over another range of prices (see Figure 4). 8m (3. the induced demand (under the regime of q = b . The generic linear demand curve is given by Q = A − mp . p2* = . and ( p2 . The seller wants to realize ( p1. We now determine the maximum revenue and the optimal values of p1 and p2 under Group-Buying. and the seller has to set prices before the demand regime is realized. q1 ) when the realized regime is q = b . He offers prices of p1 and p2 for quantities demanded up to the thresholds q1 and q2 respectively. Let the two curves be given by: m q = a − m1 p and q = b − m2 p where a>b and m1 ≥ m2 .  2m  a 2 + mb 2 The seller’s expected revenue is π p1* . where q1 ≤ q2 and p1 ≥ p2 . This should be 2m seen as a preventive measure to stop the market from clearing at a sub-optimal price. if quantity demanded q ≤ b − 2m . At a price of p2 . We normalize m2 to 1 m2 so that the two demand curves are given by q = a − mp and q = b − p . as seen in Figure 4. the profit maximizing Group-Buy price-quantity schedule is given by: a b  p1*   2 .mp. the two curves intersect only when b > . If the demand regime given by b − p is realized.  Price P* =  *  =  (3. The seller chooses the quantity b − a to separate the two prices. [INSERT FIGURE 4 HERE] Figure 4 shows a case of intersecting demand curves. Lemma 3: For the model of intersecting demand regimes described above. Further. This price-separation quantity can be explained as follows. the price could slide down to p2 when the optimal price is the higher p1 (see figure 4).p.p ) is 14 . and we m assume that this condition holds. q2 ) when the realized regime is q = a . otherwise. Each demand regime can occur with equal likelihood. where a>b and a m ≥ 1 .3) ( ) Remarks: 1. We derive the optimal revenues from Group-Buying and posted price markets and compare the two. Under Group-Buying. the seller picks a pricequantity schedule as before.

Thus. Thus the choice of q2 ' = b − as the quantity-discount 2m 2m threshold preserves the profit maximizing prices under both regimes for the seller. unlike the earlier case of parallel demand curves (wherein the two demand regimes had identical slopes). 15 . when the seller relies on posted prices. the single quoted price is ineffective in exploiting the demand heterogeneity. a 2. and compare these with the optimal prices and revenues under the Group-Buying mechanism. given by q2 ' = b − p2 = b − We now derive the seller’s optimal price and revenues under posted pricing. it is easier to induce a self. If the demand regime given by a − mp is realized. the revenue to and posted pricing is given by π g − π p = 8m ( m + 1) the seller from Group-Buying strictly dominates the revenue from posted pricing. 17 See Tirole (1989). * * p = and π ( p ) = 2 ( m + 1) 8 ( m + 1) (ii) The difference in revenues between the optimal schedules for Group-Buying 2 ( a − bm ) > 0. However when the seller specifies that the 2m minimum quantity demanded (fo r a price of p2 to be offered) should be greater a than b − it prevents the market from clearing at a price of p2 under this 2m demand regime. Group-Buying does better under greater demand heterogeneity because it enables the seller to set (non. and we can easily check that 2m 2 a a * ' q2 > q 2 = b − .a . thus maximizing his total expected revenues. (i) The profit maximizing price and revenues under posted pricing are given by 2 a+b ( a + b) . Group-Buying outperforms posted prices for intersecting demand curves. via quantity-discount schemes. wherein the two demand regimes have different slopes. With greater heterogeneity in demand regimes. in the following proposition. the seller is forced to make a trade-off between revenues in one demand regime and revenues in the other. Here. then the choice of q2 ' = b − 2m as the quantity-discount threshold will not affect the market-clearing price since a a * the demand induced by p2* = is q2 = . Thus. However.linear) price-quantity schedules that optimize revenues under each demand regime. Proposition 2: In the case of intersecting demand regimes.selective seconddegree 17 price-discrimination among customers.

The posted price mechanism makes this tradeoff in the case of parallel demand curves as well. for all m. i. As this happens. Hence the quantity-price coordinate at which the two demand curves intersect. GroupBuying might outperform simple pricing. 5. p ) =  . posted prices perform as well as GroupBuying in the earlier case. As the relative-slope parameter m increases. Thus. A related issue is the absence of scale 16 .e. We expect. the  m −1 m −1 dominance of the demand curve q = b − p over the other ( q = a − mp ) increases. Remarks: 1. Since the heterogeneity in the two demand regimes stems from the difference in the m slopes of the two demand curves. the commitment to production quantities precedes the pricing decision. this asymmetry of dominance translates of course to greater demand heterogeneity.e. this curve rotates downwards counter-clockwise (see Figure 4).  . Perhaps the seller is unable to exploit the richer space of price schedules under Group-Buying (compared to posted prices). Note that the demand curve given by q = a − mp is anchored at the point ( a. but since Group-Buying is also ineffective in enforcing price-discrimination (due to the homogeneity of the demand regimes). Group-Buying schemes become more attractive relative to posted pricing... Pricing and Production: Timing and Scale Economies Our previous models implicitly assumed a certain sequence of events. with greater freedom in the timing of the pricing decision. When pricing follows production. as the asymmetry of dominance between the two demand regimes increases (i. there are fewer degrees of freedom in the pricing decision. For example. therefore. while demand homogeneity would neutralize the advantage of pricing flexibility (enabling second-degree price discrimination) that Group-Buying pricing mechanisms offer. that as the heterogeneity of the demand regimes increases. the assumption of zero marginal costs is reasonable when production costs are sunk. given by  a − bm a − b  ( q. Put differently: under a different sequence of events. Proposition 3 below establishes that this is indeed the case. under both Group-buying and posted prices. Group-Buying would become more attractive relative to posted pricing.Making this tradeoff leads to lower overall expected revenues as well. Proposition 3: The gains from using the Group-Buying mechanism increase as the demand heterogeneity (m) increases. because of the limited flexibility in pricing. 0) on the quantity-axis. moves down and to the right. and m = 1 measures this difference. demand becomes more heterogeneous). we expect m2 that the gains from Group-Buying increase in m.

The valuation of each type is drawn uniformly from the unit interval [0. (i) what is the impact of economies (or diseconomies) of scale. under either market mechanism. when the firm determines the price or priceschedule. the supplier commits to the total production quantity before the revelation of demand through the price (or priceschedule). Our interest. It is evident from inspection that this does not alter our results in any fashion. when the firm has to commit ex ante to purchase or production quantities. is in comparing the relative performance of the two market mechanisms (posted prices and Group-Buying) under each case. In a model such as ours. This is feasible when production or procurement lead-times are sufficiently small.e. and now needs to determine the optimal price (in the case of posted prices) or price-quantity schedule (in the case of Group-Buying). and 17 . We also assume that both the supplier and the buyers are risk-neutral. the firm will do better when pricing precedes production/procurement (which is a form of production postponement). one customer type has a higher value than the other with probability 1. It may choose to sell less than the entire quantity available. Each type of customer is homogeneous in valuation. it has no control over production quantities. As N is a multiplicative factor that characterizes all revenue expressions under both mechanisms and since we are interested in a comparative ranking of the two mechanisms we will omit N from further analysis. to see if the order of production and pricing affects the choice of the market mechanism employed. Scenario 2: Pricing precedes production / procurement: This corresponds to the case where the firm has the option of deciding on the production quantity after determining the quantity demanded for its price or price-schedule.. Case 1a: Pricing follows production / procurement under constant marginal production costs: We assume that there are two types of buyers (customers) in the market. Thus. We first analyze scenario 1. since Group-Buying is an extension of traditional quantity discounting. this is common knowledge. Thus. and buys at most one unit of the good.economies in our previous models. We assume that the supplier is a monopolist.1]. i. and (ii) does the order of production versus pricing matter? We analyze two different scenarios in this Section. and N customers of each type. Production is ex ante. under the additional assumption of constant marginal costs. its greatest benefit might be precisely to maximally exploit scale economies. where the uncertainty is on the demand side rather than in production. Scenario 1: Pricing follows production / procurement: This corresponds to the case where the firm has already committed to production or procurement. than when it has to commit early to production quantities. however. After all. This scenario is realistic particularly under long production or procurement lead-times. This Section throws the spotlight on these two issues: In the relative performance of Group-Buying versus posted prices.

We will first analyze the case of constant marginal production cost: we normalize the marginal cost to 0. We now extend the previous analysis to the case of production scale economies for the supplier. The most general price schedule under Group-Buying for our problem may be represented as [(p1 .maximize their expected profits (value).q1 ). We model the discovery of the optimal Group-Buy pricing schedule as a game between the monopolist and the buyers. quantity discounting (and. further lowering prices. Recall that. a customer will buy the good if her value is greater than the price. Proposition 4: When pricing follows production. a customer may bid even when the current price under Group-Buying is greater than her value for the good-. Kohli et al. p2]. which translates to a unit price of p1 if the quantity demanded is ≤ q1 . Thus.5 p (ii) Group-Buying: P* = P2*g = 0. it is possible for the supplier to offer 0.5 and π * = 0. The production 18 . Myerson (1991)). in determining her own bid. (i) Posted Pricing: Pp* = 0. and a unit price of p2 otherwise— the quantity discount is thus (p1 . However. To solve the Group-Buying problem. Lee et al. Hence. Monahan (1984).1 or 2 units under each market mechanism. Buyers then choose to buy or abstain.this may induce other customers to bid. the unique subgame-perfect equilibrium and supplier profits under posted pricing and under Group-Buying are as given below. we employ the Revelation Principle (cf. an argument can be made that our assumption of constant marginal costs is suspect: as the academic literature previously surveyed (cf. Group-Buying) is particularly effective under scale economies of production or procurement. The monopolist ‘folds’ the expected resulting demand into the stage game(s) to arrive at the optimal price-quantity schedule.5 1g g Thus the equilibrium solution as well as the supplier’s profit are identical under GroupBuying and posted pricing. each customer needs to factor in the probability of bidding by other customers (which is a function of their values). (1989)) indicates. with only one difference: the supplier’s production costs are given by c(q). The monopolist determines his production quantity. depending on which of these is optimal. under GroupBuying. Case 1b: Pricing follows production / procurement under production scale economies: The model analyzed here is identical to that of the previous case (with identical game structure). In the case of posted prices. by extension.p2 ). The following Theorem derives the unique (subgame-perfect) equilibrium for this game. Thus. Then he picks a price or price-schedule. Since production precedes pricing. (1986). under both posted prices and Group-Buying. The sequence of events is as follows. the best that the supplier can do under Group-Buying with constant marginal costs is to mimic the optimal posted pricing scheme— offering quantity discounts does not improve the supplier’s profits. each customer’s bid creates positive externalities for all other customers (by lowering the expected price). A customer will buy (bid) if her value is greater than the expected price.5 and π * = 0.

The altered sequence of events is as follows.the monopolist determines 18 Assuming that c(⋅) is increasing in q. the primary difference is that the sequence of events is slightly altered. Thus.induction . but once decided upon.5 Supplier Profits for q=2 π * = 0. and then analyze the case of scale economies. While the cost structure determines the optimal production quantity (0. and the supplier the n tailors his production to exactly satisfy his commitments (orders). production costs are sunk for the pricing problem. 19 . Case 2a: Pricing precedes production / procurement under constant marginal production costs: This model is almost identical to that of Case 1a. In this case too the optimal price schedule is discovered by backward. buyers respond with their purchase decision (demand is realized). 1 or 2) and profits under each mechanism. one could argue that since production costs are ‘sunk’ before making the pricing decision. surprisingly. We do not impose any restrictions on c(q).5 1g π * = 0. procurement) scale economies. Group-Buying does not do better than posted prices under every arbitrary cost-structure. More specifically. We initially focus on the case of constant marginal costs. would Group-Buying outperform simple posted prices? This question is addressed below by reversing the sequence of production and pricing. its advantage (if any) does not derive from production (or.5 − c ( 2 ) g Table 1: Solution and Profits for Case 1b As Table 1 shows.5 − c ( 2 ) p P* = P2*g = 0. As before we model the discovery of the optimal Group-Buy pricing schedule as a game between the monopolist and the buyers. although we would expect that c(0) = 0.) is increasing in q. and c(.quantity is a decision variable. Market Mechanism Posted Prices GroupBuying Equilibrium Solution for q=1 1 Pp* = 3 1 P* = 1g 3 Supplier Profits for q=1 2 π* = − c (1) p 3 3 2 * πg = − c (1) 3 3 Equilibrium Solution for q=2 Pp* = 0. for each possible production quantity (q = 1 or 2). However. the preceding model does not provide adequate scope for the firm to exploit the generality of the Group-Buying mechanism. The supplier first quotes a pricing schedule. the equilibrium solution and supplier profits are identical under both posted prices and Group-Buying. if production decisions could be made contingent on the information revealed by the pricing mechanism. although Group-Buying is an extension of traditional quantity discounts with coordination among multiple buyers. and a convex c(q) would result in diseconomies of scale. In fact. a concave c(q) would correspond to scale economies. 18 The table below summarizes the results of the analysis under both pricing mechanisms. the maximum profits under GroupBuying are achieved by mimicking the posted price mechanism thereby resulting in a single price equilibrium even under the Group-Buying scheme.

Thus.linear) price schedule of Group-Buying. The following Theorem derives the unique (subgame-perfect) equilibrium for this game.independent fixed costs. where 0<c<1.the optimal price-schedule that will lead to the optimal revenues in the terminal stage of the game. Finally we analyze the more complex case of scale economies when pricing precedes production in the sequence of events discussed below.linear) Group-Buying price schedule can perhaps be better exploited. i. to extract and take advantage of demand information. under both posted prices and GroupBuying. without loss of generality. 20 . the richness of the (non. the unique subgame-perfect equilibrium and supplier profits under posted pricing and under Group-Buying are as given below. the supplier in our setting does not garner higher profits from the ability to tailor production to the demand revealed by the complex (non. 1g g 2 2 The production quantities under each pricing scheme are determined by the realized orders. The actual production quantities under each pricing scheme are determined by the realized orders. Group-Buying does not provide any exploitable informational advantage over simple posted pricing. production scale economies are induced by declining marginal production costs. in this case also. there are no production. for i = 1. i. Proposition 5: When pricing precedes production.. We assume that the cost of producing i units is ci . 1+ c (1 − c )2 (iii) Posted Pricing: Pp* = and π * = . we set c0 = 0 . Case 2b: Pricing precedes production / procurement under production scale economies: We now analyze the case where (i) there are production scale economies and (ii) the sequence of events is such that pricing precedes production. p 2 2 1+ c (1 − c) 2 (iv) Group-Buying: P* = P2*g = and π * = . the marginal cost of producing the second unit (given by c2 − c1 ) is not greater than the marginal cost of the first unit (given by c1 − c0 = c1 ). In our setting. Thus the equilibrium solution as well as the supplier’s profit are identical under Group-Buying and posted pricing.e. Certainly. We also assume that the marginal production cost is c per unit. We see that. As before we model the discovery of the optimal pricing schedule as a game between the monopolist supplier and the buyers.2 .e. Such production postponement is feasible when production lead-times are short enough and/or customers’ willingness to wait is long enough. c2 − c1 ≤ c1 . Since the supplier now has more options..

Proposition 6: Under production scale economies. Similarly. [INSERT FIGURE 5 HERE] The difference in seller profits under Group-Buying and posted pricing when production costs change is a function of two factors: (i) Absolute profitability. As Figure 5 illustrates. because the cost of the first production unit is c1 and the cost of the second production unit is c2 − c1 . when pricing precedes production. setting c2 = 2 ⋅ c1 in Proposition 6 leads to ( c2 − c1 ) = c1 . 1 − c1 + c2 (1 − c1 ) 2 and π * = . i. is measured by the concavity of costs.the two market mechanisms yield identical profits to the seller in this limiting case alone. the level of profits under both Group-Buying and Posted Pricing fall. which under either market mechanism is clearly a decreasing function of production costs. ceteris paribus. To illustrate. This is easily seen. (ii) Scale economies. the second unit costs 21 . A drop in absolute profitability would tend to drive the profit difference down as well..e. the profits for the monopolist from Group-Buying dominate those under posted-pricing.The following Proposition derives the unique (subgame-perfect) equilibrium for this game. Thus. Introducing scale economies while holding every other feature of the model constant resulted in Group-Buying dominating Posted Pricing. the marginal production cost of the first and second units cost are identical. as c1 increases. as is obvious by comparing the results of our analysis of cases 2a and 2b (Propositions 5 and 6). which also play an important role. in the absence of fixed costs. under both posted prices and Group-Buying.e. Plugging these values into the profit expressions yields the profit expressions of Proposition 5-. the degree of economies achieved due to scale (which. how steeply the marginal costs of production fall) is increasing in c1 and decreasing in c2 . the first unit costs more to produce and the second costs less—the relative production cost of the second unit versus the first also falls.. the unique subgame-perfect equilibrium and supplier profits under posted pricing and under Group-Buying are as given below. Further. p 2 − 2c1 + c2 2 − 2c1 + 2c2 1 + c1 (ii) Group-Buying: p1* = and 2 6 (1 − c1 + c2 )  1 * p2 =  4 + c2 − − 1 +3 ( 2 − c1 ) c1 + ( c2 − 4) c2  and 6 2 − 2c1 + c2  3/2 1 2 2 3 π GB = 52 − 72c1 + 36 c1 − 15c2 − 18c1c2 + 9c12c2 + 12c2 − 2c2 + 2 (1 +3 ( 2 − c1 ) c1 + ( c2 − 4 ) c2 ) 108 (i) Posted Pricing: p * = ( ) We find that in this (final) case. i. Group-Buying does provide an exploitable informational advantage over simple posted pricing.when costs ( c1 or c2 or both) increase. as c2 increases.

the only cost that matters is 1 * * c 1 . It turns out that P ≥ P* ≥ P2 in all cases. with the equality holding exactly when 1 c2 = 2 ⋅ c1 . Thus. and the production cost of the second unit versus the first rises.more to produce while the first unit costs the same. is varied along its entire feasible range. Since the probability of one customer having a high valuation and the other having a low valuation is high. A more interesting case is when c1 increases: also from the preceding discussion. Both P and P2 are increasing in c2 . Overall. it becomes clear that when c2 increases. as seen in Figure 5. 1 1 Under posted pricing. But then. and when both customers have low valuations. this tradeoff is reflected in the optimal posted price P* . P1 is independent of c2 . illustrating the significance of production scale economies when pricing precedes production. while the scale economies driver will cause the profit difference π g − π p to increase in c1 . From the preceding discussion.5 ). the cost-savings from avoiding production pre-commitment means that these losses are less than they would have been if production commitments had to be made before orders are revealed. Intuitively. But these events have a relatively low probability. the direction of change of the profit difference when c1 increases depends on which of the two drivers dominates. and eventually converge to P* when c2 = 2 ⋅ c1 . Group-Buying has its drawbacks-. the difference in monopoly profits between Group-Buying and posted pricing is an increasing function of c1 for the entire feasible range of c1 (and c2 ). c2 . the impact of scale economies driver dominates the absolute profitability driver for the entire feasible range of c1 . 22 .work in the same direction and drive down the profit difference. This risk is offset by the safety net of a lower P2* . the cumulative production cost for two units. The cost parameter for the first unit. Group-Buying affords the monopolist greater flexibility. to make the sales when both customers have moderate valuations.when both customers have high valuations.1 ). is set to one of three values: Low ( c1 = 0. the seller cares about the value of P only when exactly one unit is sold. Of course. Moderate ( c1 = 0. For each fixed c 1 .25 ) and High ( c1 = 0. the seller sets P* to be high to trap the high value customer’s demand when one 1 customer has a high valuation and the other has a very low valuation. c 1 . leaving money on the table. Figure 5 shows that. since this is the 1 only scenario under which P affects his revenues. but an increasing function of c 1 . the seller has to balance out the benefit of sales at a higher price (in the event of high customer valuations) versus the risk of losing sales at the higher price (in the event of low customer valuations). Furthermore. both of the above drivers. in fact. [INSERT FIGURES 6 & 7 HERE] Figures 6 and 7 shed further light on the behavior of prices and profits under scale economies under the two mechanisms. the absolute profitability driver will cause the profit difference π g − π p to fall.absolute profitability and scale economies. surprisingly. Figure 7 shows that for any fixed level of c 1 . no sales are made. the price settles down to P2* . In fact. Figure 6 plots the optimal prices under Group-Buying and Posted Prices.

and eventually becomes zero when c2 = 2 ⋅ c1 . Information and Valuation Revision In this final Section. 6. V + θ } to {0. Under the low impact scenario. we model both scenarios as being equally likely. we model a different kind of uncertainty where buyers revise their valuation of a product based on information received about the quality and attributes of the product. the extent to which information affects buyer valuations (in either direction) is given by the parameter θ. we normalize the values {V-θ. with variance in the sellers’ reputation for service and reliability. Of the four cases analyzed above. Some revise their valuation upward while others revise their valuation downward. they revise their priors (valuations) either upwards or downwards. Without loss of generality. accurate order fulfillment etc.com and Amazon). Information about the reliability of the seller (reputation for prompt service. this is the only one in which Group-Buying outperforms posted pricing. buyers are less affected by the information. which we call high and low impact scenarios respectively. Once buyers process this information. Information can induce customer heterogeneity. For expositional ease. 23 . such upward revision of priors (by a fraction f of all buyers) results in a valuation of V + θ while downward revision results in a valuation V-θ. when production can be tailored to meet revealed demand. There are two possible states of impact. We assume there are n buyers all of whom enter the market with identical initial valuations of a product given by a valuation V . In the next Section.the profit difference is greatest for c2 ≈ c1 (maximum scale economies). scale economies allow the exploitation of non. Under the high impact scenario. it falls as c2 increases. the extent of impact is given by λ. The sellers of the product provide information about product features and quality ratings.linear price schedule) and the optimal exploitation of that information (via sequencing production after pricing) together play a role in driving our results. and may affect the relative performance of the two price mechanisms. We call this ratio f the valuation revision ratio. Buyers process this information resulting in a change in their valuation of the product. The impact of the information in either (high or low impact) scenario results in a fraction f of buyers revising their valuation upwards while the remaining buyers revise it downwards. Online reviews of products consist of both positive and negative information about the product.1} respectively. Under the low impact scenario.linear pricing under Group-Buying. To summarize. As before they demand a single unit of an atomic good. Under the high impact scenario.) and the experience of other buyers with the product is available at the sites of third party infomediaries (such as Epinions. revelation of demand information (through a non. When products are characterized by complexity. We model an additional layer of complexity in that the magnitude of the impact of the information – on buyer willingness to pay is not known to the seller. we throw the spotlight on the role of product information as it affects the buyer directly (and the seller indirectly). Clearly. buyers may rely on the information provided by other buyers and product reviews available in online communities [Solomon (1999)].

the following are the optimal prices and the resulting seller revenues: 1 Case 1: x ≤ 2(1 + 2f)  p1*  1. where λ ≤ θ.x. θ ) .as before. otherwise. Note that higher values of the impact symmetry ratio correspond to lower variance in information impact. If x ≥ Lemma 5: Under the Group-Buy scheme. and corresponding seller revenues.f 1  1  * then P* =  + x  and π = fn + x  . where 0 ≤ x ≤ . If x ≤ 1. and vice-versa. p2* =  + f − x  .f n 1 1  * then P* =  − x  and π = (1 + f )( − x) 2 + 6f 2 2 2  1. Hence we refer to x as the impact symmetry ratio. a higher θ value of x corresponds to greater symmetry of impact between the high and lo w impact scenarios ( λ .. Lemma 4: In a posted price market.e. under posted prices and Group-Buying.and Revenue. 2 + 6f 2  2  2. 2  2  ( ) Finally. Since x is a measure of the  2 2 2  λ ratio of the two magnitudes of impact (under the two scenarios).  n 1 and Revenue. while a lower value corresponds to greater asymmetry of impact (i. + x  . if quantity demanded q ≤ f ⋅ n. λ = θ ) in the two scenarios. 1 1 1  this results in values:  − x. the following are the optimal prices and the resulting seller revenues: 1. q2}. and λ ≤ θ. they revise their valuations to V + λ and V.λ. When normalized.   2 2  1 Case 2: x > 2(1 + 2f) 1  1  * * Price P* =  + x  ∀ {q1 .  Price P* =  *  =  1 p   2   2 . The following two Lemmas derive the optimal prices. π * p1*. we compare the seller’s revenues from each of these mechanisms in the following Proposition. π ( P ) = f ⋅ n  + x  . 24 .

e. this result is independent of other parameters of the model. and (ii) high variance of information impact. and. 25 . Thus. these are characterized by medium to high information complexity [Notess (2000)]. Group-Buying schemes are no more attractive than fixed pricing schemes. Simon(1991)] for further reading in this context.. As buyers' willingness to pay increases. over the past few months? Our results shed some light on the reasons. we know that under high impactsymmetry.e. Furthermore. to make the purchase decision [Appelman (2001)]. buyers operating under ‘information overload’ (i. [INSERT FIGURE 8 HERE] Figure 8 shows that Group-Buying dominates posted pricing in markets characterized by a lower buyer willingness-to-pay (where the valuation revision ratio f is relatively low19 ). From our previous analysis. The items offered on sale on the B2C sites discussed in our introductory remarks as well as in the Trade journals cited were mostly consumer durable goods. This leads to the well-known herding effect. to high impact-symmetry (low variation in information impact). In this 2 (1 + 2 f ) case.Proposition 7: (i) The revenues to the seller from Group-Buying dominate the revenues from 1 posted price market when x ≤ and 2 (1 + 2 f ) 1 Revenues from the two mechanisms are equal when x > . Group-Buying emerges as the dominant option only when the product is characterized by (i) low willingness-to-pay on the part of the buyers. (ii) Thus. in effect. For such goods. This leads us to the important question: Why did the high-profile electronic GroupBuying sites aimed at consumer aggregation fail without exception. or the ratings of other buyers and online communities. Group-Buying cannot do better than mimicking the posted price mechanism. as the impact symmetry ratio x increases (i.. Group-Buying would dominate simple posted pricing only for products for 19 f is low only a small fraction of buyers revise their priors upwards while a large fraction (given by 1 − f revise it downwards. overwhelmed by the combination of information complexity and bounded rationality ) often use only a subset of the information available to them20 [Lissack (1997)]. Recall that if 20 The reader is referred to [Simon (1979) . Further. Group-Buying ceases to be the sole dominant mechanism even under low values of the valuation revision ratio. Proposition 7 shows that the relationship between the two parameters—the valuation revision ratio f and the impact symmetry ratio x — determines whether GroupBuying does better than posted pricing or not. Figure 8 illustrates this result graphically. variance of information impact falls).

our survey of the popular Group-Buy sites revealed that low-end (and low margin) consumer durables such as telephones. relatively unprofitable products: this business model was unsustainable. sunglasses and Cassette Decks were the most popular items sold. While our analysis establishes that a monopolist cannot do better than a posted price market with the Group-Buy scheme. Thus. In reality. before he can consummate the sales to customers. high end durables such as luxury coffeemakers. in practice Group-Buy markets are likely to underperform posted price markets. he cannot be certain whether the price will fall to p2 or remain at p1 . The many examples from the trade and business presses of failures of GroupBuying schemes confirm our findings. Group-Buying therefore entails a waiting period for the buyers during which the seller has not determined the equilibrium price. 7. while readily available through a variety of fixed-price sites. the best that the Group-Buy scheme can do is to mimic posted prices. He has to wait a while before he is able to ascertain the realized demand schedule and the price 21 . CD players. multiple price points) by using quantity-based discounts. In fact. Bose Lifestyle Home Theatre systems.e. This delay and the resulting price-uncertainty often lead to utility decay for the buyer.margin. buyers’ willingness to pay for these products tend to be lower than in a posted price market where the transaction is instantaneously completed. when the seller manages to induce differential pricing (i. Group-Buy schemes often do not provide higher revenues to the seller. resulting from its operational dynamics. the Group-Buy scheme is often dominated by simple pricing. Group-Buy sites found themselves squeezed into categories of low. For instance when a quantity q < q1 is being demanded. high-end Nikon SLR cameras and top-of-the-line Canon camcorders were conspicuously missing on the major Group-Buy sites. so that the seller’s revenues are the same under Group-Buying and posted pricing. In contrast. In our analysis. In such cases. In fact.which buyers have a low willingness-to-pay. due to coordination. For the seller to be sure that he can offer a price 26 . waiting and uncertainty related costs. we expect to find that the Group-Buy mechanisms under-perform posted price markets.. We explored how sellers may use Group-Buy mechanisms to respond to uncertainty in demand. Some of the inefficiencies associated with the Group-buy scheme are as follows: A seller has to be certain about which demand regime (and corresponding price) has been realized. Our model showed that under conditions of demand uncertainty that commonly obtain. Conclusion and Extensions We began by investigating some of the reasons for operating Group-Buy pricing schemes. 21 p2 he has to wait until there is a build up in the quantity demanded so that it exceeds q1 (where q1 = al − p1 or q 2 = a h − p2 : q2 > q1 ). we had not considered the erosion in buyer’s valuation of the product from the inefficiencies surrounding the Group-Buying process. As a result. and since buyers can anticipate the delay (as was the case on Mercata and MobShop) this results in an ex ante lower willingness-to-pay on her part and thereby shifts the demand curve downward.

and does as well (by simply mimicking the fixed pricing scheme. It is enough if they are alike.e. Secondly. As the demand regimes become more and more similar. made the market it itself unattractive from the standpoint of profitability. When the distribution that characterizes buyer valuations is known beforehand. the range wherein Group-Buy does strictly better than posted prices (low values of both the Valuation Revision ratio and the Impact Symmetry ratio) is also characterized by low buyer willingness-to-pay. Thus. sellers are almost always better off by running a posted price market—in fact. even though they may have different offsets. given the utility decay experienced by buyers due to the nature of the market mechanism. both buyers and suppliers face increased uncertainty about the final clearing price. the demand regimes do not need to be identical. the entire Group-Buying 22 For the advantage of Group-Buying to vanish. However. Firstly. the seller is able to use GroupBuying to price-discriminate and capture some of the revenues lost by setting a single price across both demand regimes. since their bids need to be based on the expected price rather than the actual price. it becomes even clearer that buyers that will be attracted to Group-Buying site will be those that are price sensitive. this business model could not be sustained. There are practical difficulties with Group-Buy schemes that we have not modeled here explicitly but which can make the working of the scheme more inefficient. for a wide range of parameter values. when the stochasticity with respect to demand is generated because buyers update their prior valuations after consuming information about the product. as for instance if their slopes are the same. Group-Buying does offer some advantages over posted pricing. We find that the value of Group-Buying as a price discovery mechanism depends on the nature of the uncertainty about buyer valuations in the market. the only exception was when pricing (and order revelation) preceded production and there were scale economies. sellers that operated Group-Buying sites were impacted by two factors: the products that were sold on these sites were such that Group-Buying gave no real advantage to the seller and when Group-Buying did emerge as the dominant market mechanism. resulting in unusual and dissimilar demand regimes.like low. Group-Buying strictly outperforms posted prices for the seller. Further.When buyers are heterogeneous.margin products such as the cheaper consumer durables. Group-Buy sites found themselves cornered into selling commodity. In the presence of both demand uncertainty and heterogeneous demand regimes. Unfortunately.linear pricing schedule afforded by Group-Buying enables precise exploitation of consumer heterogeneity through self-selectively induced differential pricing. Here the non. As a result. This when combined with the fact that GroupBuying is a dominant scheme only when a higher proportion of the buyers have lower willingness to pay. the situation begins to resemble the cases analyzed in Proposition 1: the advantage of Group-Buying over posted pricing shrinks and finally vanishes 22 . and providing no volume discounts) in the remaining range. results in a market whose revenue potential is considerably diminished. in such of these markets the seller was hit by adverse selection . Over the longer horizon. the conditions that made Group-Buying the dominant mechanism. This is particularly a problem when customers are risk-averse. 27 .i.

Typically. For example. Infomediaries — online companies like CNet and PDA Buzz that trade in information –attempt to establish the primacy of the information they provide. The industry. 28 . explicitly or implicitly. There are other possible approaches to effectively employing web-based market mechanisms. sub-optimal profits to the supplier. it might be an effective auxiliary scheme when used in tandem with other market mechanisms such as posted prices and auctions. perhaps deterred by the initial wave of setbacks. their bounded rationality might lead to sub-optimal behavior-. It is important to point out that even if Group-Buying is not efficient as a stand-alone approach. the market mechanism itself could be an effective price discrimination tool. customers need to make. Information has become increasingly recognized as a valuable commodity and a strategic tool. (This was reflected in our solution methodology. Future research should study the effect of these factors on the effectiveness of complex market mechanisms such as Group-Buying vis-àvis simple posted prices. Hence it is important for manufacturers and distributors to control the dissemination of product-based information. more importantly. When buyers can get signals about product quality from the bids made by other buyers (similar to common values auctions). Another possible extension is the analysis of Group-Buy mechanisms when there is uncertainty about the quality of the product. fairly complex calculations in order to optimize their bidding thresholds. Manufacturers and distributors can employ GroupBuying and other complex market-mechanisms as a strategic response to reduce competition and counter the commoditization of products induced by simple posted prices in competitive markets. or when transactions cease and the price stabilizes. which finessed this problem by employing the revelation principle. in part by commoditizing the physical products. particularly for web-based selling. Thus. the sales and clearing prices are finalized on a specific day committed to by the supplier.) Customers might experience a disutility from this buying process.and hence. customers have to incur the costs of delay in closing the transaction. GroupBuying mechanisms can be a powerful way for the seller to induce buyers to signal product quality information to each other. Thirdly. has not so far experimented with such combinations. and to monitor the progress of the Group-Buying constantly.process typically takes from 10 days to a month. in order to make their bids at the right time and price.

Management Science.. Shankland.. Marschak. 17. “Elements for a Theory of Teams”.com/news/0-1007-200-4372403. Oxford University Press. Third Edition. D. "I Scream You Scream: Consumers Vent over the Net". EContent.html 3. Sep 1986. “A Quantity Discount pricing model to increase Vendor profits”. 205 ..com/news/0-1003-200-1851571. 25-44.Whinston and J.A. March 4. J. 1991. No: 6. H. "Of Chaos And Complexity: Managerial Insights From A New Science". Economic Theory of Teams. Management Decision. 20.2. (Sep. Hau L. Irwin Series in Production Operations Management. Marschak. Jean..R. Simon. 1999. 1997. 1979). A. pp. 13.References: 1. J. 32.R. 9. Princeton University Press. Vol. “A Cooperative Game Theory Model Of Quantity Discounts”. K. pp 1177-1185.P. June 1984. The Theory of Industrial Organization. Lee. 2. Nahmias.com/news/0-1007-201-1518733-0. Game Theory: Analysis of Conflict. 14.com/awdaily/dailynews/january01/1-011501. M. 8.. Appelman... Monahan. Value America Story: http://news. No: 9. Steve. 12.. April 2000. Jun 1989. Tirole. Yale University Press.. Myerson. The MIT Press.. “A Generalized Quantity Discount Pricing Model to Increase Supplier's Profits”.. 4. The Industry Standard. Simon.auctionwatch. and Radner. and Rosenblatt. " Rational Decision Making in Business Organizations". 30.. G.. pp. Vol.B. "Web Wanderings: Consumers' Revenge: Online Product Reviews and Ratings". Management Science. Microeconomic Theory. Harvard University Press.A. 1995.218. pp. No. R. 18. McGraw-Hill. 7. and Heungsoo P. 11. 493-513. (Spring. 1972. 10. Notess. H. Mas-Colell. R. 5. Kreps.cnet. 69. The New York Times. G. R. pp 693-707. 35 (1997). S. Vol. Vol.D. 1955. 720-726.. Management Science 1.html 15. M. Dennehey. M. "Mercata Shuts Doors" http://news. A course in Microeconomic Theory. 5. David... 19. 6. October 11. 1991). M.. Vol. Lissack. " Customer Reviews: It's All a Matter of Opinion". No.cnet. J. Kawamoto. Production and Operations Analysis. "Sun Micosystems Tries Group-buying Strategy" http://news. The American Economic Review. 4. " MobShop Closes Consumer Division" http://www. No:6. 1989. Sandoval. Management Science. 35. 2001. 2. " The Journal of Economic Perspectives. Kohli.html 29 .html 16. Solomon.cnet. 1990. pp 127-137.Green. H.

com GBP: The Group Buying Partnership .uk/products.letsbuyit.htm IRPG: Independent Restaurant Purchasing Group: http://www.com: http://www.jsp MPLY: McNopoly.Group-Buying Markets Cited in Section 3: Group-Buying: The Current State of Praxis APPA: The Maryland Public Service Commission: http://www.md-electric -info.bazaare.independentrestaurants.com/ COL: Chennai Online Bazaar (India): http://chennaionline.http://www.md-electric -info.happymany.com/ MPSC: Maryland Public Service Commission: http://www.http://www.htm TBG: The Buying Group .com/home/default. 30 .groupbuying.com/ ECON: e.e.com/infocenter/aggregators.co.pine.org/MS/groupbuying.html STBZ: The StockBuzz Market: www.mcnopoly.conomy.com Vide Appendix-I for a more detailed list of Group-Buying markets.com/index_en.com THM: The Happy Many: The Happy Many: http://www.the-buying-group.buildbyte.Conomy – operated by PriceWaterhouse Coopers: www.asp PINE: The Printing Industries of New Englnad: http://www.onlinechoice.html BBYT: BuildByte Construction Portal [Indian]: http://www.com/index.com/ LBI: Let’s Buy It.com – http://www.stockbuz.html OLC: Online Choice: Online Choice: http://www.com/lbisite/index.

com HCIS Group Buying: http://www. The Buying Group: http://www.onlinechoice.Com [Egyptian Group-Buying Market]: http://groupbuy.com The Happy Many: http://www.org/groupbuying.com/ Ciao.com/index_en.Com (UK.htm Group Buying Partnership (UK): http://www.Com: (Thailand): www.bazaare.mcnopoly.html From Egypt.asp McNopoly.groupbuying.htm (B2B Health care services).com/ Business-to-Business Group-Buying Markets Independent Restaurant Purchasing Group (US): http://www.economy.html.contractsxml. Germany.Appendix – I: Group-Buying Markets Business-to-Consumer Markets in the US: Online Choice: http://www.letsbuyit.uk/ StockBuzz.independentrestaurants.com/kategorien/1.Conomy – PreiceWaterhouseCoopers: www.com – (A German Group-Buying directory): http://www.com: http://www.com/\ Chennai Online Bazaar (India): http://chennaionline. The Printing Industries of New Englnad: http://www.happymany.htm 31 .html Clust.202643.hcis.ciao.clust. France): http://www.com/ MaxGroup (US): http://www.org/MS/groupbuying.Com [French Consumer Market]: http://www.fromegypt.asp IP Circles: http://www.maxgroup.pgca.htm Printing and Graphics Communications Association: http://www.com/ e.com/Services/b2b_fremont.the-buying-group.com/ LetsBuyit.org/pages/groupbuy.com/users/Default.stockbuz.pine.com/home/default.55164.org/ecap2000/students/FINAL/f-markonew/page3.218630.co.

com/cgi-bin/search/index. Master Source Corp. The Center for Non-Profits: http://www.Printing and Imaging Association of Mid-America: http://www.html 2.appanet.njnonprofits.org/groupbuy.php BuildByte Construction Portal [Indian]: http://www.shopmates.com/ BazaarE.piamidam.Group buying services for businesses (and consumers): http://www.org/about/why/aggregation/f_buyingpower. The Maryland Public Service Commission: http://www.html 4.cgi?ID=981843428 Group-Buying Practices in Non-Profit Organizations 1.buildbyte.com/ 32 .pdf 3.mastersourcecorp.md-electric -info. APPA: American Public Power and Environmental Agency http://www.com/ ShopeMates [Group-Buying Enabler]: http://www.com/index. : http://www.com .rekha.org/groupbuy.

Appendix 2: Figures and Diagrams Figure 1: Two Demand Regimes aH aL q = aH .p Quantity aL aH High and Low Demand Regimes Figure 2: Higher Price Point From High Demand Regime And Lower Price Point From Low Demand Regime.p Price q = aL . aH aL P1 Price P2 q'1 q1 q2 q'2 Quantity aL aH 33 .

0) Quantity 34 .mp q1 q'2 q 2 (b. aH aL P1 Price P2 q1 q'2 q2 aL aH Quantity Figure 4: Intersecting Demand Curves: Neither Demand Regime Dominates The Other Universally. (0.Figure 3: Higher Price Point From High Demand Regime And Lower Price Point From Low Demand Regime.b) q=b-p (0.0) (a.a/m) Price P1 P2 q = a .

Also observe that the difference π GB − π PP is increasing in C1 and decreasing in C2 .πPP) 0.5 0.6 0.1 0. Difference in monopoly Profits between GroupBuying and posted pricing (πGB.4 0.075 0.05 0.8 0.4 C2 C1 0.7 0.7 0.5 35 .6 0.8 0.025 0.Figure 5: Profit comparisons when Pricing precedes production under Scale Economies: Group Buying (finally) does better than posted pricing for the seller.

3 C2 0.Figure 6: Price comparisons under varying costs: Figure 6 consists of three graphs that plot prices for varying costs.1 ).625 0.8 C2 (c) High Costs (C1 = 0.57 P2 P* P2 0.55 0. P1 and P2 are the prices under the optimal Group-Buying schedule.5 0.55 Prices P* 0.4 0.50 Prices P* 0.12 0.7 P1 Prices 0.5 0.51 0. P1 0.1) 0.25 ) and High ( c1 = 0.7 0.16 (a) Low Costs (C1 = 0.25) 1 23 This is the range of values for c2 that is feasible under production scale economies. while P* is the optimal posted price.5) (b) Moderate Costs (C = 0.62 0.52 0. Moderate ( c1 = 0. 36 .2 C2 P1 0. c 1 takes one of three values depending on the cost regime Low ( c1 = 0.5 ) while c2 varies23 between c1 and 2c1 .6 0.53 P2 0.

7 0.405 0.25) c2 c2 (c) High Costs (C1 = 0.30 0.20 0.16 0.31 Profits Profits 0.Low ( c1 = 0.Figure 7: Profit comparisons under varying costs: c 1 takes one of three values depending on the cost regime .12 0.1 ).1) Profits 0. 0.2 c2 0.18 0.415 0.19 0.425 (a) Low Costs (C 1 = 0.14 0.25 0.24 0.45 0.8 (b) Moderate Costs (C1 = 0. 37 This is the range of values for .5 0.25 ) and High ( c1 = 0.22 0.35 0.23 0.5) Group-Buying 24 Posted Pricing c2 that is feasible under production scale economies.5 ).5 0. Moderate ( c1 = 0.6 0.32 0. while c2 varies 24 between c1 and 2c1 .4 0.21 0.30 0.29 0.18 0.

4 0.8 1 Group-Buying Dominates Group-Buying & Posted Price Markets are Identical Impact Symmetry Ratio Valuation Revision Ratio Comparison of Seller's Revenues Under The Two Market Mechanisms 38 .3 0.25 0.5 0.4 0.35 0.2 0. 0.6 0.2 0.45 0.Figure 8: The two markets compared in terms of the two ratios that result due to buyers’ revising their willingness to pay.