You are on page 1of 17

Course Points

&vic-yo March

ismJon -> Find to pic


you March Submission + Presentations
-

[Find exam
open
book) X No spss
Scanning of Market
-

USTRAA CFEMALE RAZON]

IOL:-Attitude,
A Interest, opinions, Lifestyle

Collect information [scon the environment]


see
1) me internal record [For company having past datal
cookies.
REM
a MIS

Recency, Frequency, Monilization

2) Market
intelligence systems -
Data that is available outside
↳Distributors / Intermediaries at a
very cheap price
-
Sales Force
·
Mystery Shoppers -Available market
helps in
doing
>Buying competitor products demand forecasting
↳>
Review $ Raling sites -

Target market is used to do


> Live
Agencies sales
forecasting
↳Government Data calculated
-

Potential market is

by doing market research

Market
3)
Research

Market Potential -
Arggty x Price x Potential

purchased No. A
buyers
-
Potential Market v

V Internally Most
Available Market you can get Important
-
data
Target Market
Lakshman Food India Limited Case
MD ) MDZ
-
What is the
management dilemma ? Mkt share
,
How to differentiate

What decision should management take ?


-

Market research problem What significant attributes ?


-

are

what information is needed ?


-

MIS DIAGRAM

Secondary Data

Internal External

1
CRM Data Warehouses Syndicated Govt sources
-
Non -
Govt

Social Media

Management Dilemma Market Research

*
Ask what the decision -

Info needed Break MD into

maker needs Focus underlying small parts which


-

on causes

Focus
symptoms Info oriented basically gives
-

you
on
my
-

-
Action oriented your MR .
MARILET RESEARM
DATA

Secondary Primary RESEARCH DESIGN

2ua Even

EXPLORATORY CONCLUSIVE
n.y R.D
v

Trying to understand
~

insights Direct
Small sample size
Descriptive Causal
Surveys -Experiments
-

Interviews FGD
otone Cross sechina

Voltas Longitudine
Line

Indirect
Use Coses SProjective techniques
17 Understand user
V
I
perception/preferences Aid 70

2) Impression of Word Assi Sentence Completion Sentence Expression startthe

new
product/price Construction interviews (FGDs
Mode of Admin
3] Observation >
Story board High unconvering
-

on
/ Personal
mechanical
Structured Unstructured Ardit subconcious mind
Disguised Undisguised Content
Trace
Analysis
Selective Perception [Researcher's Bias]
-

Investigator Triangulation
·

Westerndrop Price model (price for


qua
-

Van
sensitivity serves as
proxy
SUVIDHA TRUCKS LP: HOW TO CONDUCT A FIELD EXPERIMENT

INTRODUCTION

One day in early August 2017, Jerry Singh was walking around his employer’s parking lot. Singh was the
research director at Suvidha Trucks LP (STLP), a company based outside Toronto that owned and
operated over 300 “big-rig” long-haul trucks. Part of his job was to use his modest research budget to help
improve the fuel efficiency of the company’s fleet. Because of the scale of the company’s operations,
even a tiny improvement in mileage could translate into massive savings on an annual basis.

Earlier that morning, one of his colleagues had shown him a brochure for a fuel additive that had just been
introduced to the market. Called Optimum, it was manufactured by a U.S. firm that touted the product’s
fuel economy benefits. The advertising suggested that Optimum offered superior results, but Singh was not
convinced. Every year, he saw or heard about dozens of technologies and products that purported to offer
similar benefits, but they nearly always failed to deliver. Optimum, though, seemed more credible than
most because it was endorsed by a large trucking company in the southern United States that appeared to
have no business links to Optimum. Was it right for his Canadian operations too?

Singh had to make his recommendations to the company’s top management in April 2018. He knew that
Baljeet Jenkinson, STLP’s chief executive officer (CEO), would reject any recommendation that was not
supported by evidence. How could Singh determine if Optimum was mere hype or a way to help STLP’s
bottom line? His job depended on it.

SUVIDHATRUCKS LP COMPANY BACKGROUND

STLP, a large Canadian transportation company established in 1994, had a number of specialties, including
the transportation of oversized goods, solid and liquid waste, shipping containers, automobiles, hazardous
goods and industrial machinery. While the company had operations throughout North America, the
majority of its activities were in Canada and the northeastern United States. In a typical year, its trucks
travelled more than 100 million kilometres (km) and transported millions of tons of goods. About half the
trucks operating under its banner were company-owned while the other half were owned and operated by
subcontractors. Regardless of who was behind the wheel, everyone who worked for STLP was dedicated to
providing excellent customer service at affordable rates.

OPTIMUM FUEL ADDITIVE PRODUCT INFORMATION

Optimum was a new fuel additive designed to improve fuel economy by burning fuel more completely
without compromising oil life or emissions. It could be used in any gasoline or diesel engine and contained
no hazardous or toxic elements. The product retailed for about $40/bottle and each bottle was enough to
treat approximately 330 gallons (1,250 litres) of fuel (see Exhibit 1 for a conversion chart).

Singh found an online discussion group of long-haul truckers and read a thread focused on Optimum. The
anecdotal evidence was mixed. While some drivers reported no positive results, others were more
optimistic:

I started using Optimum back in 2016. Since then, I have driven 10,200 miles. I am driving a
2010 Freightliner Columbian with a 470 horse Detroit engine and a 13-speed transmission, and to
my delight, I’ve gone from 7.5 to 8.1 mpg [miles per gallon].
TEST: BIG RIG

After learning what he could about Optimum online, Singh decided he needed to get some first-hand
experience with the product, so he asked some of his drivers to start using it and keep records of fuel
consumption. He would have preferred a large test, but that would be too expensive and time-consuming.
He settled on a 30-day test that would use six relatively new trucks and involve only experienced drivers
who were likely to take the test seriously (i.e., accurately record mileage, etc.).

Since STLP kept careful records of the fuel mileage achieved by each of its trucks, Singh planned to
compare the fuel mileage results of these six test trucks against historical data he had for similar trucks
currently operating in the fleet.

Singh also made what he thought was a clever choice: while the drivers he chose were mixed in terms of
using company-owned and owner-operated trucks, they all drove the same route. That is, all the trucks
drove the “garbage run” between Toronto and a landfill outside of Windsor, Ontario (about 200 miles or
330 km away). This meant all the trucks would run full in one direction (to the landfill) and return empty
on the same road at about the same time of day and in the same weather conditions. In fact, all the truck
drivers involved started the study on the same day, with each driver subsequently recording between 20
and 27 data points 1 for each truck. Singh also provided each of the drivers with instructions on how to use
the product.
After a month, Singh looked at the results (Exhibit 2). He was not sure what to conclude. The historical
fuel mileage achieved by similar trucks in the STLP fleet was typically just under five mpg and the test
trucks had beaten that, recording 5.5 mpg. This was good news. On the other hand, a visual inspection of
the results revealed a disturbing pattern. It appeared that several trucks had experienced a decrease in
mileage over the test and there was a lot more variance in the results than he had expected. After spending
a lot of time trying to understand these results, he realized that he had little confidence in the test.

LOOKING AHEAD

Singh was aware of the importance of coming up with a solution to increase fuel savings for STLP, and
the quarterly business review meeting was fast approaching. Now, he was trying to figure out how to
proceed: if fuel savings and increased gas mileage were apparent, he would have a much stronger position
making the case for introducing Optimum fuel additive to the fleet.

With the confusing results of the study and not much time to explore alternatives properly, Singh was
preparing his final report. In the end, he had to make an “up or down” recommendation: should STLP
start using Optimum more widely?

1
During the test, some drivers took holidays while others took extra shifts. This explains why not all drivers have the
same number of observations.
EXHIBIT 1: GALLONS TO LITRES, MILES TO KILOMETRES CONVERSION CHART

Gallons = Litres Miles = Kilometres


1 = 3.8 1 = 1.6
2 = 7.6 2 = 3.2
3 = 11.4 3 = 4.8
4 = 15.1 4 = 6.4
5 = 18.9 5 = 8.0
6 = 22.7 6 = 9.7
7 = 26.5 7 = 11.3
8 = 30.3 8 = 12.9
9 = 34.1 9 = 14.5
10 = 37.9 10 = 16.1
15 = 56.8 11 = 17.7
20 = 75.7 12 = 19.3
25 = 94.6 13 = 20.9
30 = 113.6 14 = 22.5
35 = 132.5 15 = 24.1
40 = 151.4 16 = 25.7
45 = 170.3 17 = 27.4
50 = 189.3 18 = 29.0
55 = 208.2 19 = 30.6
60 = 227.1 20 = 32.2
65 = 246.1 25 = 40.2
70 = 265.0 30 = 48.3
75 = 283.9 35 = 56.3
80 = 302.8 40 = 64.4
85 = 321.8 45 = 72.4
90 = 340.7 50 = 80.5
95 = 359.6
100 = 378.5
EXHIBIT 2: TRUCK PERFORMANCE FOR SIX TRUCKS, STUDY 1

Truck 1 Truck 2

Truck 3 Truck 4

Truck 5 Truck 6

Note: Achieved miles/gallon over number of observations. Higher values on the vertical axis indicate better fuel efficiency.

Source: Case authors.


Open coding
Content analysis Disclosure analysis Conversation analysis
Axial coding
Narrative analysis Qual DA software NVIVO
Themes
Solve this
Selection bias, systematic error Internal validity and External validity<
using randomisation

Experimental Designs
/
X = TREATMENT
• One shot study x O1
O = SINGLE OBSERVATION OF DV
• One group pre test and post test
DV = DEPENDENT VARIABLE
O1 x O2
IV = INDEPENDENT VARIABLE
• Static group design
EG = EXPERIMENTAL GROUP
EG x O1
CG = CONTROL GROUP
CG x O2
• True experimental design
EG: O1 x O2
CG: O3 x O3
1.
Specify the information needed

of method
2.
Specify the type interviewing
3. Individual content
question
4. Overcome the unwillingness to answer

4.
Who will be
responding
4.2 Filter questions
4.3 can the
respondent remember [Telescoping]
4.4 Reduce the efforts
4.5Sensitive Information

5.
Choosing structures [structured/ Unstructured]
a
question

mca Dichotomous Scale


Shuffle
* the options

6.
Choosing question wording
6.1 Use simple/unambigous words
62 Avoid leading questions
7
Determining the order of questions
ended in middle of
7.1 Avoid putting open question
Survey
7.2 Putdifficult end
at
questions
7.3
Move, from general to
specific
8.
Pretesting -
Test on a small scale

8.1 Debrief candidate


your

SPSS

-Empirical analysis oftransfer of brand doyalty in

brand octcnsious
Scaling is actension of measurement
-

Nominal
-
Ordina

-
Interval
-
Rotio
> [Description]
It for test
Nominal is only description. No statistical can
-

be performed on this.

> [Description Order)+

Ordinal There of
is order
objects.
-

an

[Ranking)
The of distance is
quantity some.
not

Interval -
Lorder Distance
+

Description]
+

Ratio -

[Description - Order -
Distance Origin]
+

Saching Techniques

(Non-metric) Comparative [matric Non Comparative

Constant Continous Likert


Semantic Stapel
Paired Differentic <10
(Summated)
-

Sum (Bipolar) point


Rank 2-sort
Used in OB sccle Scala)
7
objects
sort some criteria
W Order
on

and rank them


-
Non-metric
W
-

mesured
Several
in relative > Sum total of Cre No V
objects are
neut rol
terms asked
allocated units point
~

Conjoint Analysis
Points of caution: -

1. Number of scale
categories: - Don'tgo beyond T

2. Balance and unbalanced scale: Balanced scale means

to have neutral option and unbalanced scale has


no neutral option.

IndependentSample T-Test Coding /Editing


Paired Sample T- Test V

Cleaning
Leave the missing value as blank or

99
puti tas

Test
Levene's of
homogenity
p 0.05or
= less -> EU not assumed
p> 0.05 -> EV Assumed

T- Test: -
Difference between two
groups in terms of brand

loyalty

TURF
-BL

PVP 7 PL

>S2

Cronback 4 0.7

Narrow
Cigar - linear relationship

Ranked data ->


Spearman
Correlation
-

N
~

Direction between variable [Given by Sign)


-

Strength of Relationship -> Chan (1988)


Coefficient of determination to 0.29 Small
-

r 0.10
=
nour

media
=
r 0.30 to 0.49 corr

v 0.5
=
to 1 darge corr

Regression
Multiple R
-

-
Mierarchical
Wise R
Step
-
Sampling Design
-
who do
you want to talk to? -

Defining TargetPopulation
-

Sampling frome

Sampling Procedures [Sampling Techniques]


-

(Randomness)
Probability (more Non-Probability
ructure)
sampling Sampling
-

Every member of
-

Probability of
the
population has member
any being
known non-zero
selected is unknown
of
probability
selection
Convenience
Quota

Judgement Snow Ball


2
most easily
V
available
Experien und
29. Intervice individual > Dividing
into
on Road
selects people
population smaller

-
used in am basis of groups.
exploratory personal judgement Es. Which age group pratirs
research
which brand of cigrette
in
stratified choosing
a
particular city?
quota randomly
-

simple Random
Sampling
Systematic Chances of
getting systematic
-

error

Cluster We
-

look at
a instead individuals
city of

I
* STATETHETECUNLIVES IN YOUR PROJECT

mDs [Multi Dimensional Attribute


Scaling)...
-


Based
Conjoint man
[Multi Discriminent]
-Wilk's Landa of
Test -> Between
->
equality 8-1

Smaller the attribute,


value ofthe more
significantit is

-Wilk's Landa table to find how dimensions to


many
select
by looking at
significance value

whom

against what value


For or TG market, JGneeds
where

your brand
name, reason

now ↓
why TG

should believe

your POD

You might also like