The Ultimate Guide To The Evolution of Programmatic Buying AdClarity E Book

THE ULTIMATE GUIDE TO
THE EVOLUTION
OF PROGRAMMATIC BUYING
AdClarity
MARKETING INTELLIGENCE
Welcome to the Ultimate Guide to the
Evolution of Programmatic Buying!
In this e-book, you’ll learn what

programmatic buying is and what
components or key players are
necessary in order for programmatic
to exist. Your journey will begin
with the traditional media buying
landscape and the transitions it
took in order for programmatic to
appear. You’ll learn about
everything from the key players to
the challenges facing the industry
today and how Marketing
Intelligence is the key to
programmatic.
We hope you enjoy your read

and look forward to answering
any questions you may have.
www.adclarity.com 2
TABLE OF CONTENTS
Chapter 1 Chapter 6
Traditional Ways Programmatic Direct
Chapter 2 Chapter 7
DSPs and DMPs Mobile Advertising
Chapter 3 Chapter 8
SSPs Obstacles & Challenges
Chapter 4 Chapter 9
Ad Exchanges Marketing Intelligence
for Programatic
Chapter 5
RTB Conclusion
www.adclarity.com 3
CHAPTER 1
TRADITIONAL WAYS
Advertisers and Publishers
In the old times if an advertiser wanted to buy inventory from a
publisher, they would do so manually and directly. This was a
cumbersome and tedious process which involved negotiations,
insertion orders, manual tracking, and long waiting periods
(especially for premium sites). The benefits were so-so as the
advertiser had to track the result of each site manually in order to see
where they were performing the best. Additionally, because
advertisers bought on a CPM model, they bought impressions in bulk.
Regardless who the user was, everyone was seeing the same ad.
It was literallywho
Regardless like a Jackson
the user Pollack painting- finding the canvas you
want
was,toeveryone
work on and
was just throwing all your paint on it, hoping that at
seeing
least
the one drop
same ad.reaches the right spot.
It was literally
like a Jackson Pollack
painting- finding the
canvas you want to work
on and just throwing all
your paint on it, hoping that
at least one drop reaches
the right spot.
www.adclarity.com 2 4
TRADITIONAL WAYS
As time went on and the number of advertisers and publishers grew

exponentially, it became impossible for publishers to sell most of their inventory
and for advertisers to handle the direct media buying process. They would have
to be in direct contact with thousands of publishers at a time in order to
maximize their reach. Additionally, manually tracking all of these sites proved to
be impossible. This issue paved the way for the ad network.
Ad Networks
An ad network is a company that connects advertisers to a large amount of
publishers. It acts as a type of broker between the publisher and advertiser,
buying the unsold or remnant inventory of the publisher in order to sell it to the
advertiser. The ad network had technology that enabled them to categorize the
inventory, which they could package and sell to the advertiser. This benefitted
the advertiser because it allowed for them to extend their reach to numerous
publishers all at once and have their ads targeted a little better based on the
categorization of the ad network. And rather than having to track the
campaigns on the advertisers end, the ad network was now responsible for
tracking and optimizing each campaign across multiple sites.
And history repeated itself.
As time went on, the number of ad networks grew and began flooding the
market. This put the publishers and advertisers in another predicament. Now
they had to spend time trying to figure out which ad networks would give them
the best bang for their buck. It also meant that advertisers could be buying the
same inventory on the same site more than once. Oh, and did I mention that the
ad network was taking a chunk of the ad profit from the advertiser?
And thus, the Ad Exchange was born.
TRADITIONAL WAYS
Simply put, an ad exchange is a technology platform that facilitates in the

automated buying and selling of inventory, often through real-time
auctions. Unlike an ad network, purchasing is not based on a CPM model,
but rather per impression. An ad exchange is pretty much a giant pool of
impressions that allows for advertisers to bid on each impression based
on specific targeting criteria in real-time. This allows for the advertiser to
make sure each ad they are serving is being served to the right person at
the right place at the right time. With the ad exchange, the publisher is
now able to gain more revenue by selling off more inventory, and the
advertiser has greater transparency and control over their placement.
And although ad exchanges made things better, advertisers and

publishers still didn’t have the tools and technology needed to really
exhaust the full value and potential of ad exchanges.
So along came Demand-Side Platforms (DSPs) and Supply-Side

Platforms (SSPs). Both of these platforms are very similar, but just serve
different audiences. The DSP serves advertisers and the SSP serves
publishers.
DSP
A DSP is sort of the brain on the advertiser side in terms of the ad
exchange. It is a platform for buyers which works with multiple ad
exchanges, allowing them to buy from multiple sources of inventory. For
the advertiser, it is the interface which allows them to manage their bids,
create targeting criteria, aggregate all their user siloed data, do
retargeting, optimize their campaigns in real-time, and have access to
TRADITIONAL WAYS
results and data from DMPs. It could be seen as a sort of ad

network on crack without the huge mark-up and with automated
technology capabilities. To sum it up, a DSP is a one-stop shop for
advertisers; all they need to do is submit their ads and wait for the
results.
SSP
An SSP is pretty much the equivalent but for the publisher side.
Whereas the DSP’s goal is to buy ad impressions as cheaply and
quickly as possible from the ad exchange, the SSP’s goal is to
maximize the price of their impressions. The SSP’s enable
publishers to have access to multiple ad exchanges, ad networks,
and DSPs all at once, increasing the range of potential buyers.
So now you understand the key players in the programmatic buying

ecosystem. And all programmatic buying is, is the automated
process of buying or selling media. And this process is reliant on
these key players in order for it to work.
CHAPTER 2
DSPS AND DMPS

We’ve established the very basic components that are necessary
for the existence of programmatic buying. But we need to
understand a few more in order to understand why things work
the way they do and how they work the way they do. This chapter
will specifically focus on the core and functionality of the DSP.
In the previous chapter, I mentioned that the DSP acts

as the brain, while the ad exchange acts as the body.
But the brain is composed of cells and neurons, both
critical to the functionality of the brain. In this case,
we’re interested in the thousand billion neurons that
provide us with the capacity to reason, comprehend,
and make decisions.1 So whereas the DSP acts as
the brain, it still needs these neurons in order to have
a purpose. And in this case, data management
platforms (DMPs), are the DSPs’ neurons—they give
the DSP access to tons of information and data.
DSPS AND DMPS
So what is a DMP?
A DMP is a very intelligent technological platform that takes data from first,
second, and third part data providers2 . It manages the data in an organized
and meaningful manner so that the DSP can make the best purchase
decisions possible on behalf of the advertiser.
First Party Data

Considered to be the most powerful and effective of the three types
of data, first party data is owned data, which means that the data is
based on what you, as the advertiser, has collected on your
audience. It is directed at site visitors or existing customers and is
usually accessed through pixels, tags, cookies, and can also include
information from CRM systems or web analytics. It is a crucial
component to unique advertising and retargeting which allows
advertisers to leverage internal behavioral data to meet their goals.3
Second Party Data

This is the newest form of data which definitely has the potential to
be the most powerful once it reaches its maximum capabilities. All
second party data is, is somebody else’s first party data. It’s usually
made available through an arrangement with partners or companies
who have the same type of target audience. For example, a luxury
female clothing company might partner with a luxury jewelry
company in order to broaden their advertising reach. The consumer
behavior and demographics should overlap, which would allow for
the jewelry company to target the female clothing company’s
audience and vice versa.
DSPS AND DMPS
Third Party Data

Third party data is the monster of all the data providers. It is
pretty much all the data that isn’t in the realm of first party
data. It’s not connected to the advertiser and is an
independent party which gathers data from a multitude of
websites, vendors, and, even offline. It can be melded with
first party data to provide an even more targeted campaign or
be used on its own. Third party data provides the advertiser
with additional consumer behavior that is unattainable
through first party methods. It allows them to paint a more
colorful picture of a user in order to give them the best fitted
campaign based on their behavior.
Example 2 Combining third party data with first party data
Someone came to your site to buy a TV and you placed cookies on

their browser to track them and retarget them. But what are you going
to retarget them with? With the help of third party data, you will be
able to gain a more in-depth look at their behavior. For example, what
if you knew they were also in the market for a Porsche? You would
have more information about their income, type of lifestyle, type of
image they would like to portray, and so forth.4 This could reinforce
your decision to show them an ad, not just of a generic 42” TV, but of
the most progressive high-tech luxurious TV out there on the market
(also the most expensive).
DSPS AND DMPS
Example 2 Using isolated third party data
You’re a sewing machine company who wants to get new customers.

You can find this consumer through third party data by seeing if the
consumer had visited any sites related to your target audience, such as
a sewing magazine or crafts website. Regardless if they visit a photo
website or sports website after, you will be able to show your ad on the
site because you know that they had recently visited a site that was
correlated with your target audience.
So now we know the first step- how the DMP gets and imports its data.
What next?
Now the DMP needs to segment the data and normalize the data. All
this means is that it has to find trends, and segment your audience
based on consumer behavior (i.e. women in California between the ages
of 18-27 who use iPads, etc.).
Next, the DMP combines step 1 (the accumulation of data) and step 2
(the segmentation of data) to send instructions to the DSP. The DSP
then makes the decision on how much that impression is worth for the
advertiser. And of course, the impression comes from the publisher’s
side, or the SSP. But we’ll get to that step in the next chapter. For now,
you’ve grasped a basic understanding of how the DSP functions. Next,
you’ll see how the SSP functions and where the two meet.
CHAPTER 3
SSPS
This chapter covers the supply-side platform, or SSP. In layman’s
terms, an SSP is the publisher-side equivalent of a DSP whose
main goal is to maximize revenue for the publisher and, well,
make their life a whole hell of a lot easier.
How?
In today’s online advertising ecosystem, there are a plethora of ad networks,
DSPs, and ad exchanges. And if the publisher wanted to sell their inventory to
them, without an SSP, it would be close to impossible to do this efficiently and
effectively. Dealing with every platform individually would be an extremely
manual and labor intensive process.
Ad Network
Ad Network
DSP
Advertiser
Publisher
DSP
Ad Exchange
SSPS
It would take months to do any sort of A/B testing to see who is bringing in the
best results, require a ton of time from the media seller to negotiate, test, and
run every single deal, and, it would be almost impossible to optimize their
inventory as the market is so volatile, results that are not based on real-time are
irrelevant. Oh, and did I mention that if they did want to do any sort of
automatic optimization they would have to integrate some very heavy and
complex technology into their server?
So the SSP comes along and says, “Hey! We can not only manage your entire
inventory in an automated manner, but we can also bring you premium traffic,
lots of traffic, take all manual labor off the table, bring you the technology you
need, and, make you money!”
Ad Network
Ad Network
SSP DSP
Publisher Advertiser
DSP
Ad Exchange
SSPS
So now, all the publisher has to do is provide the inventory for the SSP and the
SSP does the rest of the work. And this is where it gets very, very cool.
You see, there is a special and complex algorithm consisting of three main
variables embedded in the SSP which decides who gets to win the impression.
Think of it as a sort of multiple regression analysis and multiple correlation (R).
Uhh… WTF is a regression analysis?
A regression analysis is simply a way of using one variable, or an input,

to predict an outcome. In a multiple regression analysis you use several
variables to predict an outcome. And usually, after you run your
regression analysis, you run a multiple correlation to see which variable
has the strongest relationship, or correlation, with the output.
You have three variables and you want to see which variable has the highest
correlation with the output, or in this case, the maximum revenue. And every
single time an impression is put up for auction, the results change based on the
values of the variables.
The three variables are:

1 RTB
2 Payouts
3 DMPs
RTB
I’m not going to get into too much detail here, because I am going to explain
real-time bidding (RTB) later on in the guide. But all RTB is, is the means in
which an impression is bought and sold in real-time. This process takes place
in an auction where it is being bid on. And just like in real life, the person with
the highest bid wins the deal. So in advertising terms- the advertiser who
SSPS
places the highest bid on that impression wins the space to put their ad up!
I want to make a note here: the highest bid through RTB is usually the winner.
Probably 90-92% of all won impressions are based on the highest bid. However,
we will be covering the other two variables as well because they are important
and although not often, they do occur.
Payouts
Payouts usually only happen when the advertiser is buying on a CPA or CPL
basis. The advertiser usually offers to give the publisher a certain percentage of
the profit. In the case where there is a tie for the bid, the SSP will sell the
impression to the third party who has the highest payout rate. This case
happens very rarely, but the concept still exists as an option in this ecosystem.
The DMP
Have you ever been at one of those booth games at a carnival where they give
you a colored ball and you have to throw that colored ball in a matching colored
cup? Easy enough. You know what the color is that you need to target based
on the color of the ball.
But what if you didn’t know the color of your ball and you had to throw it in the
correct cup? You would be aiming at literally nothing, with the hope that you
may have gotten the ball in the right cup. So in order to win, you would HAVE to
know what the color of that ball is.
SSPS
This is the role that the DMP plays with the SSP. The DMP in this case
provides the SSP with behavior data about a certain user and “colors” or
segments them. Now that the user’s color is known, the SSP knows where to
direct their traffic to in order to increase the likelihood of success.
Confused?
Let me draw it out for you.
The SSP has contact with multiple ad exchanges, DSPs, and ad networks. It
wants to get the ad that has the highest likelihood for a conversion. Without
the information from the DMP, it wouldn’t be able to know how to segment the
user and would just throw the impression at any DSP, ad exchange, or ad
network. But with the information on the user, it would know which mediator
would be the most applicable for this inventory to increase the likelihood of
conversion.
Phew!
Ok, so you now understand the three major inputs that manipulate who wins
the impression. And you also know that in most of these cases RTB wins.
To sum it up, the SSP is a platform which helps publishers sell their ads to
huge pools of buyers, more than they would have on their own. They do so
primarily through RTB so that the publisher can get the most value and
money for each piece of inventory.
CHAPTER 4
AD EXCHANGES
We already know how and why the ad network came to be. In
short, there were so many advertisers and publishers in the eco-
system, it was impossible for advertisers to handle the direct
media buying process and it was impossible for publishers to sell
most of their inventory. So the ad network came to be.
For advertisers, this was very good because it enabled them to extend their reach
and reach a larger number of publishers through one platform. For small and
medium publishers, this was also very good. This was because these publishers
had a very hard time selling their inventory to advertisers and now they were able
to sell their inventory as part of a larger package on behalf of the ad network.
Ok, so how does an ad network actually work?
Firstly, the ad network needed access to a huge audience. This audience was
brought by the publishers, and along with it, an abundance of consumer behavior
data. Next, the ad network would look at all the inventory it had aggregated and
they would use an algorithm to forecast how much of this inventory would be
available in the following months per each publisher. Next, they would aggregate
this data based on segments, which they are able to do with the use of the
consumer behavior data. They would then package this segmented data (based
on things such as age, gender, demographics, etc.), mark them up, and sell them
to advertisers.
It sounds fine…so what’s the problem?
AD EXCHANGES
Let’s start with the advertisers. First of all, most ad networks don’t want
advertisers to know where their ads are getting published. This is for several
reasons, the most important being that if you, as an advertiser, knew which
publishers your ad network was working with, you would be able to pass the
ad network and work directly with those publishers, eliminating the need for
the ad network. Because of this, advertisers can’t identify which inventory
works best for them. Next, because advertisers buy inventory in packages,
and these packages often include both remnant inventory and premium
inventory, advertisers can’t really know what the value of any single
impression is and can’t really obtain too many insights. Additionally, because
there is almost no transparency and advertisers don’t know what sites their
ads are running on, they could be taking the risk that their ads are showing up
on nefarious sites.
Publishers didn’t have it much better. The lack of transparency also ran on the
side of the publishers. They were unable to identify which advertisers were
bringing in the best results and had to take the chance that the ad network
would bring in very low quality ads, which could in turn lower the reputation of
the publisher’s site, decrease any likelihood of conversions, and so forth.
Because publishers were working with multiple ad networks at once, they had
to constantly be evaluating which ad network was bringing them the best
results—this was a manual and time consuming process. And most
importantly, ad networks, more often than not, were also working with a chain
of other ad networks, buying and selling inventory from one another. Each one
of these took a cut from the publisher leading the publisher to receive lower
and lower revenue.
And let’s not forget the challenges the ad networks themselves had. Although
their plan seemed to be seamless, let’s remember that there is no such thing
as concrete forecasting. It is a prediction, not an accurate science. So of
course, forecasting publisher inventory was often faulty. And because of this
AD EXCHANGES
instability, their packages were inconsistent and they were either

overselling or underselling inventory.
So you see the problem. What started off as a solution turned out to be a
little more detrimental to its key players than intended. And this is where
ad exchanges came into the picture.
Ad Exchanges
An ad exchange is a platform which facilitates the buying and selling of
inventory through real-time auctions. Unlike ad networks, which don’t
work in real-time and sell inventory by the bulk, ad exchanges buy and
sell per impression. This gives the advertiser complete control over each
ad- they have specific criteria they specify ahead of time so that each
impression is bought with a real purpose.
It eliminates the traditional methods that provoked wasted impressions.

Now, they are able to choose exactly who would see their ad, where they
would see it, how they would see it, and when they would see it. So not
only would they get to choose each and every impression they are
buying, but they also decide how much they want to pay for each
impression and only pay for the impressions that they want.
Ad exchanges conduct auctions for every single impression and just like
in real life, the highest bidder wins the prize. So now, rather than the
publisher selling 10,000 impressions in one bulk package and not
necessarily getting the best bang for their buck, each impression is
auctioned off to the highest bidder, which ensures that the publisher is
getting the maximum profit based on market values.
AD EXCHANGES
The key differences between ad exchanges and ad networks

Although I assume most of you are starting to understand the key differences
between ad networks and ad exchanges, it won’t hurt to reiterate them
in this section.
1
Because each impression is sold to the highest
bidder, there is high competition between the
advertisers. If they know they want it they will try to
outbid each other, raising the CPM for publishers.
2 Advertisers can now buy only the impressions

that they are interested in. Rather than buying in
bulk from ad networks, they have complete control
over their impression buys which allows them to
buy at a much higher price.
2 Transparency for all! Because ad exchanges

provide both the advertisers and publishers with
transparency and visibility, advertisers are now
able to gain insights as to which were the best
performing sites and help them optimize for future
campaigns. Publishers are able to see which
advertisers are driving the highest eCPM to their
sites, giving them the insights they need to make
better decisions regarding which advertisers they
want to encourage to their site.
AD EXCHANGES
4 If you remember, when I introduced the ad

exchanges, I said that they are a direct connection
between advertisers and publishers. Ad networks, on
the other hand, can work with multiple intermediaries,
causing the distance between the advertiser and
publishers. The problem with that scenario is that each
middleman takes a cut of the profit, lowering the
revenue left for the publisher. So ad exchanges create
an environment for a more efficient relationship,
enabling the publisher to make more money.
CHAPTER 5
RTB
Programmatic Buying
We established that in traditional media buying, the buying
and selling process was tedious, time consuming, inefficient,
not promising, and didn’t necessarily yield maximum results for
either the publisher or advertiser. And of course, everything
was and is done manually.
All programmatic buying is, is the buying and selling of display ads in an
automated manner using the technology that we covered in the previous
sections. That’s literally all it is: making media buying an automated and
much more efficient process. It’s a concept, an umbrella term that
encompasses a whole world beneath it.
You may be super confused right now and think that there is much more to
programmatic buying, which there is, but what you need to understand that
the actual definition of programmatic buying is simple. I can’t reiterate that
enough. The automated process of media buying. Period.
That being said, there are different types of programmatic buying, such as
real-time bidding, or RTB, which I’m sure you’ve all heard of. We’ll get to this
in a moment, but I’m going to drive you crazy and say this one more time:
PROGRAMMATIC BUYING IS THE AUTOMATED PROCESS

OF MEDIA BUYING USING TECHNOLOGY.
RTB
Good. Now that we have that settled, let’s begin.
People often say that programmatic buying is a synonym for RTB. And now
that you know the definition of programmatic, I am certain that you will be
more than willing to understand that it’s not--real-time bidding is a type of
programmatic buying.
So what the hell is RTB?
Real-time bidding is the buying and selling of ad space (per impression) in

real-time through auctions held in ad exchanges, DSPs, or SSPs. RTB is not
its own platform- it’s a type of technology that ad exchanges, DSPs, and
SSPs integrate in order to perform these auctions.
Okay… so how does it work?
I’m going to give you the simplest version of how the RTB works. And keep in
mind, this entire process takes less than 80 milliseconds, about 4 times faster
than it takes to blink your eye.
It all starts with the publisher. The publisher chooses an SSP to work with,
such as Pubmatic, and places a tag in the ad space, or inventory, they want
them to be in charge of.
A user goes to that publisher’s site and before the page even finishes
loading, the request to fill that vacant as space has been sent to the SSP. The
SSP finds out information about the user through its DMP. It gathers generic
information such a demographics, time of day, day of the week, and so forth.
Once all of this information is gathered, the SSP sends the request to the ad
networks, ad exchanges, and/or DSPs it is working with.
RTB
The SSPs can distribute this request in 2 ways- round robin or simultaneously.
If you look at the infographic below, you’ll see how the SSP has connections with
multiple platforms (DSPs, ad networks (AN), and ad exchanges (AE))
DSP 1
DSP 2
SSP AE 1
AE 2
AN 1
If the SSP decides to send the request in a round robin manner, that means that
each of these entities will have a ranking based on priority. For example, the SSP
could send the request to DSP1 and if the bid falls through, they’ll send it to
AN1—it really depends on what their order of preference is.
Or they can send out the request to all of their partners simultaneously.
It’s important to note this because if you are an advertiser and you are working
with a certain platform who works with a specific SSP that has traffic you want,
you need to make sure that you are not on the bottom of their list in case they do
a round robin. You need to be working with the highest ranking DSP, ad exchange,
or ad network in the eyes of the SSP.
So the SSP sends the request over to its partners. And don’t forget, if they send it
RTB
over to an ad exchange, the ad exchange could be working with several

DSPs simultaneously.
Each player on the buying side evaluates the bid according to their own
algorithms and targeting parameters that the advertiser set previously. For
example, the advertiser may have wanted to target men between the ages
of 30-47 who have a high income. Or maybe, the advertiser only wanted to
target people who surfed on luxury watch websites. Or maybe, the
advertiser only wants to target people who were shopping on their site
earlier. The possibilities can go on forever.
So if you remember, the DMPs are what provide the DSPs with behavioral
data about a consumer. The DSPs are what make the decisions. So after
the ad exchange sends the bid request to the DSP, the DSP makes a
decision based on complex algorithms which decide if this user is relevant
for the advertiser. If it is, then the DSP places a bid on the ad space (the
maximum bid amount is set previously), and whichever advertisers places
the highest bid wins the ad space. (It is important to note that not every bid
is created equally. Sometimes, an advertiser will pay more for a bid when
they know that the user is extremely relevant for them.) Then, the winner
sends instructions on how to retrieve the creative to the ad exchange,
which then sends it to the SSP, which then sends it to the publisher. And
finally, the ad is shown to the user.
Benefits of RTB
There is absolutely no dispute- RTB is a much more efficient way of buying
and selling media when compared to traditional media buying.
Advertisers no longer have to work directly with publishers and negotiate.

They no longer have to buy bundled packages with irrelevant impressions
which lead to wasted ad spend. With RTB, they are able to have enhanced
targeting capabilities, minimize wasted ad impressions, increase their reach
RTB
and frequency, have transparency to gain the insights they need to constantly
optimize their campaigns, and choose each impression they want to buy.
Publishers have the same deal. They don’t have to waste time and energy
trying to contact and manage negotiations with advertisers. They have
access to a much larger audience of potential buyers. They have a higher
likelihood to sell more of their inventory. And, most importantly, because
advertisers bid on inventory, the publisher is receiving the maximum market
value per impression.
That being said, like all things great and small, there are no ups without the
downs.
But I guess that also depends on whether the glass is

half empty or half full.
So here’s the deal.
Before programmatic came around, advertisers cared about publishers. They

cared about the publisher’s name, reputation, size, etc. Advertisers wanted to
buy inventory space because of the publisher. And because of this, the
majority of all ad spend was going to the premium publishers. No one had
any interest in working with middle or long tailed sites.
premium publishers
middle & long tailed sites middle & long tailed sites
RTB
But all of this changed with programmatic.
Rather than obsessing over the status of a publisher, buyers began to shift
their focus to the audience and this was a hard hit for premium publishers.
You see, buyers realized that it didn’t matter if they were buying premium
spots on premium sites if the user wasn’t relevant for them. All they cared
about was reaching the right user at the right place at the right time, and
buying from premium publishers just wasn’t enough.
A few things happened because of this.
First, premium publishers began to lower their inventory cost because they
couldn’t compete with natural changes. However, because the amount of
internet users increased exponentially, it sort of weighed out the situation.
Their costs may have decreased slightly, but quantity increased immensely.
Secondly, because buyers cared about the audience, this gave a chance for
the middle and long tailed publishers.
How? Through retargeting.
Buyers could now target their exact users on any site, no matter where it
stood on the totem pole, through retargeting tactics. Their specific method of
doing this is a little controversial, but it’s also pretty damn smart. It’s called
data leakage. What buyers would do is this: they would buy ad impressions
on premium sites (they would set their bid ceiling very high so they could
win). And because the ad was being served through a third party platform,
the buyer could drop a tracking code on their creative to cookie the user
(without the publisher knowing). Once the lead was being tracked, they pretty
much had no use for the premium publisher because they didn’t matter
anymore. What mattered was where the user was going, and often times
RTB
they were going to middle or long tailed sites. So the advertiser would stop
working with that premium publisher because they got the information
they need and then they would immediately launch an RTB campaign that
would look for the users with the cookies through various ad exchanges,
DSPs, and ad networks. And once they found them, they could place a
much lower bid (because it wasn’t a premium publisher) to reach them.
So on the one hand, there is a huge debate on whether RTB is hurting or

benefitting the publishers, but I strongly believe it is the latter. Yes,
premium publishers have to lower their costs and deal with data leakage.
However, with the growing number of users on the internet, they’re actually
still making a profit.
Additionally, RTB sort of leveled out the playing field. It gave a chance for
the smaller publishers, who no one would have even thought of before. The
fact that the disparities between the big and medium-small publishers is
decreasing is absolutely phenomenal and could only bring benefits to the
ecosystem.
CHAPTER 6
PROGRAMMATIC DIRECT
You may have heard nuances of this term: programmatic direct,
programmatic premium, premium direct, automated
guaranteed, programmatic guaranteed, programmatic reserved,
etc. They all mean the same thing, but for our sake, we will stay
with the term programmatic direct.
programmatic premium
programmatic guaranteed programmatic reserved
Programmatic Direct premium direct

premium direct
automated guaranteed
We have understood that with RTB, there is no guarantee that you will get
the ad spot of your choice. And we have also understood that there is no
direct relationship between the advertiser and the publisher; they are
always meeting through an ad exchange or some other mediator,
sometimes multiple mediators. Something that I didn’t mention in the last
chapter was that even though the RTB ecosystem provides transparency
into the media buying ecosystem, it doesn’t provide publishers with the
visibility they needed to understand why advertisers are buying their
inventory. It also doesn’t provide advertisers with the actual site their ad
appears on—only the category.
PROGRAMMATIC DIRECT
And we’ve also seen that in the reality of it all, although I argued that
RTB is actually really beneficial for middle and long tailed publishers, it
sort of screws premium publishers.
And the publishers are fully aware of this.
A few years ago, OpenX published a piece that included a section about
how fair programmatic buying really is. They asked some leading
publishers what they thought and most of them came to the same
conclusion: advertisers have the advantage when it comes to
programmatic.
In fact, here are some direct quotes from this paper regarding the issue:
As a premium publisher [I believe] the advantage goes to the

advertisers that are buying our inventory at below market rates.
Until the eCPM starts climbing to match direct sales eCPM, the
advantage will never be to the seller.
Buyers are “calling the shots on price, and you play the
game or don’t get the volume. Buyers are biased too much
towards audience and not enough towards context.
Buyers benefit more because programmatic “allows them to

use their first-party data to identify valuable audience and
buy it cheaply, because the seller does not know the value.
Buyers are benefiting more, as they are able to get massive

reach on top sites for bottom dollar. Meanwhile, publishers
are playing catch up, and struggling to optimize the balance
between direct sales and programmatic buying in their yield
strategies, and in particular trying to limit CPM erosion.
PROGRAMMATIC DIRECT
And of course, the advertisers had their version, but to sum it up, they
felt that the publisher side just wasn’t doing enough to fully maximize
the potential of real-time bidding.
So we see that although RTB is a stellar technology, it still has some

faults and room for improvement. And we also have to remember that
RTB is not the only way brands buy media. A lot of the top advertisers
still prefer to buy directly from their premium publishers to guarantee
their spot.
But if we remember from the first chapter in this series, the traditional
direct media buying process is lengthy, tedious, and completely
inefficient.
And this is where programmatic direct comes in.
So what is programmatic direct?
Programmatic direct is the automated process of direct media buying

where the advertiser still buys directly from the publisher, but without
the need for insertion orders AND giving them the transparency and
insights they need to constantly optimize their campaigns in real-time.
This type of programmatic means that when the advertiser buys
inventory from the publisher, they are guaranteed that spot (unlike RTB
which has no guarantee at all). And because it is programmatic and still
involved in the programmatic world, DMPs still play a huge role in
providing advertisers and publishers with the first, second, and third
party data they need to enhance their targeting capabilities.
So now, publishers are able to work directly with advertisers and see
exactly what they are looking for and why they are purchasing their
inventory. This allows for publishers to find relevant advertisers, provide
PROGRAMMATIC DIRECT
more relevant packages, and promote their own premium inventory. It’s
as if to say that now that the advertisers know exactly what they want,
publishers can provide those wants with more guaranteed inventory
than they would get in the auction place. Providing, of course, that the
advertiser is willing to pay for it.
Another common notion is that programmatic direct is only about

premium spots. But, something I strongly believe in is that premium is in
the eye of the beholder. What may be trash to one buyer may be
premium to another. It really depends on what the buyer is looking for
and what they need. So really, in programmatic direct, there is a
premium for everyone.
Let me try to break it down in simpler terms.
RTB had not yet managed to surpass the process of direct buying and
selling of inventory. Why? Because larger brands still want guaranteed
access to their premium inventory, and, well, RTB just can’t guarantee
that. So let’s say an advertiser finds the perfect customer but loses the
auction to someone who bid higher. It’s not the end of the world; the
advertiser can easily track the user to find them again. But this time,
they may not choose to go through an auction but rather do a
guaranteed buy so that they can make sure to reach them.
Additionally, when it comes down to RTB, the most important factor for
the buyer isn’t the publisher or the content on their site, it’s the audience,
pure and simple. All they care about is that they are targeting the right
audience, hopefully in a relevant manner, but not always promised.
With programmatic direct, the main factor is the actual site. Buyers want
to know that the premium inventory they are buying ad space on is
quality. Don’t get me wrong- the audience is still an extremely significant
PROGRAMMATIC DIRECT
player in the decision factor. However, programmatic direct provides the

buyer with the possibility of providing their end user with an ad that is
contextually more relevant. As an added bonus, buyers also get much
more transparency and inventory control over their purchases. And of
course, publishers are able to set their own prices, albeit much higher
than RTB prices, and make a better profit for their premium goods.
They’re also saving a ton of money on using an automated system over
manual sales work. Look, at the end of the day, you pretty much get what
you pay for.
In my opinion, programmatic direct is primarily for brands, not

performance based advertising. It’s large brands and agencies who care
about where their ads are seen. You can’t have a company like Nike
showing up on a download or porn site just because the user they wanted
is there. They have a reputation to maintain!
So how does it work?
All programmatic direct is, is the automated process of traditional direct

media buying.
The advertiser works with a publisher to buy a packaged deal of premium

spots and sites and all the IO’s, requests, briefs, etc. are done
automatically via the platform. Don’t get this confused with an ad
network, which packages inventory from multiple publishers. Rather, on a
programmatic direct platform you choose the publisher you want to buy
the inventory from. The publisher provides the advertiser with a package
of inventory only from their sites based on the general targeting attributes
that the advertiser has requested. And, most importantly, remember that
this kind of buy isn’t user based. It’s like the old ways where an ad space
bought shows up everywhere to everyone who opens the site. However, in
this case, the advertiser has limited targeting capabilities (like 30-45 year
PROGRAMMATIC DIRECT
old men, only on sports sites, etc.). So if a brand bought inventory from
Yahoo, via programmatic direct, and made a request that it would show
only to women between the ages of 20-45 and only in the fashion
category, it would show the ad to all women who fit that description in
any fashion domain or subdomain of Yahoo.
Another thing that a lot of these publishers do with programmatic direct

is offer a sort of one-stop marketing shop to the advertisers. Along with
buying inventory for their ads, they also offer packages which include
partnerships, sponsored posts, newsletter promotions, etc. If you are a
large brand who is working with CNN or the Huffington Post, this is
obviously a very appealing selling point. And of course, publishers can
make a whole lot of money from offering these solutions.
Okay. So we understood what programmatic direct is, how it works, and

why the advertisers and publishers both benefit from it.
What now?
I have an idea. How about understanding programmatic buying in the

mobile ecosystem?
CHAPTER 7
MOBILE ADVERTISING
There is no way I couldn’t include a chapter about what’s going on in
the mobile world. Unless you live under a rock, you must be aware of
how big mobile advertising is becoming. Here are just a few statis-
tics from around the web to demonstrate my point:
According to Forrester Research, mobile display advertising

will account for nearly 40% of all online display ad spending by
2019, up from 24.4% in 20145
By 2018, it is predicted that 2.7bn people on this earth will

have smartphones. Take this into consideration with the fact
that last year, 60% of time spent on the internet was spent via
smartphones (and time spent on desktops declined by 20%)6
By the end of 2015, mobile advertising in the U.S. will bring in

$400bn compared to $139bn in 20127
Okay, so I don’t need to keep reiterating the importance of mobile

marketing. We all get it. But because we’re focusing on programmatic
buying, I’m going to shove one more statistic in your face.
It is predicted that by the end of this year, mobile will surpass desktop
display ad spending in terms of programmatic ad expenditure. To be
more precise, or as precise as I can be in regards to forecasting, mobile
will account for 56.2% of the total programmatic digital display ad
spending by the end of 2015.8
MOBILE ADVERTISING
The concept of programmatic buying, as we learned in previous chapters, is

pretty much the same for the mobile world. Ad exchanges, SSPs, DSPs, etc. are
all playing the same role, except that they’re focused on mobile, and the types of
programmatic that are applicable for desktop display are still relevant for mobile.
With that being said, there are a few key distinctions that must be made in
regards to the mobile ecosystem.
1 Programmatic in the mobile world is still relatively novice. It sort

of looks like what the desktop display advertising industry was
10 years ago.
2 In the mobile landscape you have the options of choosing in-app

inventory, video inventory, or web browser inventory. Although
video is a huge deal, I’m going to focus on the two main types:
in-app and web browser
3 Native Advertising is a must.
4 When it comes to mobile, apps win.
5 When it comes to apps, Facebook wins. Big time.
6 Google and Facebook together make up almost 70% of the

mobile ad market.
7 The cookie problem. Most mobile web browsers do accept first

party cookies. However, different browsers behave differently
with third party cookies, and, most importantly, cookies don’t
exist in the app world.
8 Last, but certainly not least, is data. The types of data available
on mobile are quite different from desktop data, enabling mobile
marketers to try new targeting and optimization methods.
MOBILE ADVERTISING
Alright, let’s get started.

Mobile marketers have two main ways to advertise: mobile apps and mobile web.
Mobile Apps
Just to give you an idea of the significance of mobile apps, in 2013 80% of the
time spent on mobile devices was spent on mobile apps. In 2014, this number
increased to 86%, while mobile web browsing fell to 14%! 9 And I’ll give you one
guess as to who is dominating the mobile app industry (hint: it’s number 5).
That’s right.
Currently, Facebook is the most downloaded app in the entire world, and touts its
1.9bn monthly active users, which continues to grow each month.
So why is this interesting to us?
Because in 2012, when Facebook first launched its mobile ad business, it made
close to nothing. Yet, in 2014 it made over $12.47bn in revenue, and 69% of that
revenue was attributed to mobile.11 And even more interesting is the fact that
during the final few months of 2014, they served 65% LESS ads than the previous
year… but the average costs of those ads was 335% higher.12 Pretty much the
underlying rules for neoclassical economics. Cut supply down, demand goes
soaring, prices fly through the roof.
How did they become so successful?
There are many factors to take into consideration when it comes down to
understanding why Facebook is doing so damn well on mobile. But in my opinion,
I think the two top contributors are native mobile advertising and Facebook’s
retargeting capabilities.
MOBILE ADVERTISING
I’ll explain both of these in a little bit. Right now, we’re going to spend a
moment talking about mobile web.
Mobile Web
The mobile web is pretty much like browsing the web on your desktop, just
on your mobile. It’s just much smaller, the site needs to be mobile
responsive, and is usually touch-screen sensitive. It has quite a bit of
convenient features that enable users to perform an action in a touch of a
button (placing an order, dialing a phone number, etc.). Usually companies
always go for the mobile website first because it doesn’t require building an
app, builds up web presence, is instantly available, is compatible with
almost every mobile device, and is just much easier to manage and
maintain.
In regards to the mobile web industry, Google wins by far. However,

although Google’s mobile search ad spending is around 65% (2014), it has
been on a constant decline when compared to its 2012 figures of 83%.13
The reason for this is as discussed before, more and more people are going
to their mobile apps over web browsers. Google stays pretty private
regarding their mobile share and revenue, but what we do know is that they
are aware of this decline and probably have some tricks up their sleeve,
such as potentially acquiring InMobi, one of the biggest mobile ad networks
in the world.
The Cookie Problem

Sounds like an oxymoron, huh?
Cookies are very important in the world of advertising, especially when it
comes to retargeting. And although cookies work wonders when it comes
to desktop, it’s a little bit of a different story in the mobile world.
MOBILE ADVERTISING
APPS (WEBVIEW) MOBILE SAFARI* BROWSER/CHROME
1st Party Cookies
3rd Part Cookies
What does it mean? What does it mean? What does it mean?
Cookie support limited to Same cookie behavior as Cookie support identical

sessions in the same app. nline Safari. to most online browsers.
• Click based conversions • Click based conversions • Click based conversions
• View based conversions • View based conversions • View based conversions
YES
• Data synchs • Data synchs • Data synchs
NO
*Note that installed browsers can
LIMITED behave differently. E.g. Chrome on iOS
will support 3rd party cookies.
Source: iab
Mobile Web Cookies

In the mobile world, cookies exist ONLY on the mobile web and their reach is
very limited. And when it comes to accepting third party cookies, every
browser behaves differently. I mean, Safari doesn’t even accept third party
cookies and it’s one of the most used smartphone browsers. Two more big
issues? Browsers don’t share cookies with each other and they reset every
time you close the app session. Eeek.
Mobile App Tracking

I’m going to start off with an analogy before I start explaining the mess with
mobile app tracking.
Imagine there is a kingdom- one king to rule them all. There are tons of
people living in the village who follow and report to the rules of the king.
Things are in order. People may be living their own lives and doing their own
thing, but in regards to the important stuff (food, income, taxes, etc.),
MOBILE ADVERTISING
everything is tracked, reported, and standardized. Things function.
This analogy represents cookies and tracking in the browser world.
Now imagine the same kingdom, but this time all the villagers also play the
role of kings. Now there is one piece of land with hundreds of kings, each
setting their own rules and regulations within their own castle and property.
There is no one ruler to rule them all in this scenario- it’s literally a case where
everyone makes the rules for themselves. And because this is the case, no
one is collaborating with each other or giving other kings inside information
on what’s going on in their own kingdom.
And this is what tracking looks like in the mobile app world.
MOBILE ADVERTISING
There is no such thing as industry standards in regards to mobile app

tracking so every app is in charge of figuring it out in their own way.
Additionally, apps cannot share their tracking data with other apps so their
information stays siloed within their own app. You can imagine how difficult
this must be for advertisers when they’re trying to track their users across
different apps in order to do proper ad targeting.
Notice that I haven’t used the word cookies yet when talking about mobile
app tracking. This is because cookies don’t really exist in the app world.
Rather, there are alternate tracking methods used such as device
fingerprinting, MAC addresses, Open Device Identifier Numbers, Apple IFA
and IVA, and Android Referrer. I’m not going to get into all of these tracking
methods, but here is a really great article from Search Engine Watch that
talks about the pros and cons of each of them.
Okay, so you get the issue. Tracking is super inconsistent in the mobile
world and makes it very hard for advertisers to do ad retargeting. Marketers
must figure out a way to integrate all of these identifiers in order to have
one major identifier they can use across all channels.
But of course, there are exceptions to this problem: Facebook & Google.
Google
I don’t have much to say about Google that isn’t already known. They own
everything. And because they own everything (like DoubleClick and
Android), they have special benefits that allow them to overcome the
technical challenges that many others face in the industry.
MOBILE ADVERTISING
Facebook
Remember when I said that the two biggest contributors to helping
Facebook conquer the mobile app landscape are native advertising and
retargeting capabilities? This is where we get to it.
Website Custom Audience (WCA)

Do not get this confused with FBX. I don’t care about FBX and it’s not
relevant for the mobile world at all.
What WCA does is that it allows marketers to target their audience through
Facebook ads based on their own (the marketers) data which could have
been taken from their CRM, first party data, email database, etc. This means
that you can target people who have been on your website but aren’t your
Facebook fans or other offline customers through Facebook.
So pretty much, you take your list of customers, potential customers, and
anyone who visited your website and got cookied and upload it into
Facebook. You can segment your visitors based on which pages they
visited, what section of the website they were on, and so forth in order to
show them highly relevant ads. Facebook creates your audience based on
your list, and voila, you are able to do ad retargeting in the most extreme
and precise manner without bombarding users with emails.
And as you know, relevancy is the key to positive ROI on Facebook. The
more relevant the ad is to the user in their buying lifecycle, the more likely
they will buy!
Am I the only one flipping out about this?
This is incredible! Facebook has more data than any other data platform in
existence. It seriously knows EVERYTHING about you (probably more than
you know about yourself) and to allow advertisers to have access to this
MOBILE ADVERTISING
data through retargeting is just something else.
I think you’re starting to understand why I believe that retargeting is such

a crucial component to the success of Facebook’s mobile domination.
Native advertising is pretty easy to understand and if you really want, I

don’t mind writing about it more in detail in another article. But all you
need to know is that native advertising is really, really big (like $8bn big),
and that Facebook is really, really good at it. And once again, you
shouldn’t be surprised. Because Facebook has the data to know exactly
what they should be recommending and when they should be
recommending it.
Data
You have a general idea by now of how and where data is gathered from.
I’d like to mention one more very important characteristic unique to
mobile, and that’s location-based data.
People go with their mobile devices everywhere, and that’s something

advertisers can leverage big time. If you combine location-based data
with other types of data (such as behavioral), advertisers can send even
more hyper-targeted ads to their consumers that offer the right product
at the right place at the right time.
For example, a retailer can combine a user’s purchase history with their
location and send them personalized offers whenever a customer walks
into a store.14
It’s the ideal situation for brands that want to be involved in their
customer’s and prospect’s buyer journey.
Okay, so we understood that programmatic is pretty much the same
MOBILE ADVERTISING
when it comes to mobile, albeit in a much less mature stage. What really
shakes things up is the uniqueness of the mobile advertising industry in
and of itself. The technological limitations that mobile sets for itself
makes it difficult for programmatic to function as it does in the desktop
world.
We saw that mobile browsers and mobile apps are completely separate in
terms of data, tracking, really anything. We know that retargeting is
something that is very rare in the mobile world, unless you’re Google or
Facebook. And we know that programmatic won’t reach its full potential
unless things change.
Luckily, ad networks and other platforms are aware of this and are
beginning to develop technologies which enable them to track users from
app to app, collecting data and tracking information along the way.
And hopefully, they will be able to catch up soon. Because at the rate
mobile is growing, players can’t afford to take their time.
CHAPTER 8
OBSTACLES & CHALLENGES

Everything in life has a yin and yang, and this includes programmatic.
Yes, programmatic has made some revolutionary changes in the way
media buying is done that benefits the entire online advertising
ecosystem. But with it also comes a new type of black market where
nefarious people take advantage of human error, the rise of
automation, and the decline in human to human communication.
The most common type of fraud, and often synonymous with the
term, is non-human traffic aka bots. We will definitely be focusing on that

in this post, but I will also take note of some less popular fraud types
which are also extremely detrimental to honest players and the integrity
of the industry.
Non-Human Traffic & Bots (Click Fraud)
It’s easy to understand why non-human traffic (from this point on we’ll
call them bots) tops the list of concerns for advertisers and brands; they
currently account for 90% of all fraud in the online advertising industry.
Advertisers are paying and expecting that their ads will be viewed by
actual people who might potentially buy their product. However, this isn’t
always the case as bots account for about 11% of all display ad views.15
Bots are computer software applications driven by code which

impersonate human behavior. They are pretty advanced and can be very
hard to detect, especially as they have become better at mimicking user
behavior throughout the years.
The type of fraud that bots engage in is called click fraud, which occurs
when the bot clicks on an advertisement to drive up costs for the
advertiser without ever having the intention of conducting business with

them. Keep in mind that bots aren’t humans- they can run forever and at
an extremely high pace.
There are different types of bots such as PhantomBots, which travel

around the web viewing and clicking ads, DeceptiBots, which can mimic a
human’s behavior, and VaderBots.16
VaderBots are the most intrusive to users because they corrupt PCs and
infect them with a malware which conducts click fraud based on the
user’s behavior. There is a really great infographic that the iab made which
shows just how easy it is to become a victim of traffic fraud:17
Source: iab
Ad Injections
Ad Injections have a very interesting story.
You know when you download some file or app off the web and it makes you
“run” and “install” about a million pop-ups and have you “read” their novel on
terms and conditions (which we never do)?
Well, often times when you download apps off the internet, they will bundle
themselves into an installer which will persuade you to install additional
programs onto your computer. This is where not reading the terms and
conditions or the pop-ups and just clicking “next” “next” “next” come in. So really,
when you think you’re getting what you were looking for, you are… but you’re also
getting much more than you expected.
I don’t know if you’ve hear of this little bastard, but I’m talking about ad injectors.
You see that ad on Dell’s site that’s outlined in red? Take a moment and think
about it. Do you really think that Dell acts as a publisher? Would Nike host
ads from retailers on its site? Or Apple? I think not.
This is what ad injectors do; they inject ads onto sites without permission
and without the publisher ever knowing. And therefore, without ever paying
the publisher.
Another example
So let’s say you’re searching for boots on Google.
boots
Instantly, Google will come back with results. But in even less time, the ad
injector will push forth its own ads and replace those that paid for Google
Ok, so who gets paid here?
Normally, the money would go to Google, right? But Google isn’t making any
money here. Why? Because the ads were never put up by Google in the first
place. The ads belong to an entirely independent third party service who
hijacked the Google search page and placed their own ads as a way of
making money. So let’s reiterate this- the ads on Google have not actually
been placed there by Google!
Why do advertisers do this?
Well, this is a very attractive scheme to advertisers because now they are
being offered inventory in places they would have never been offered before!
For example, not too long ago, a Target ad actually showed up on Walmart’s
homepage. If that’s not a shocker, then I don’t know what is.
Why do Ad Injectors do this?
On one hand, ad injectors could sell inventory directly to the publishers. In

fact. If the price is good enough, advertisers may even say yes. It would
indeed be more appealing to advertisers as these ad injectors would be able
to offer up space that would otherwise be impossible to attain i.e. Walmart
or Apple. Additionally, the spots on the page (above the fold, main banner)
would definitely be an attractive aspect to advertisers.
But this isn’t what ad injectors do at all. Rather, they work through an
intricate and complex web of ad networks, exchanges and other mediators.
There are many benefits to these mediators, the strongest being the ability
to increase efficiency. However, because these webs are so complex, the

advertiser may be going through 3 or 4 mediators before getting to their
publisher. The advertiser may never even suspect that they are going
through an ad injector!
Impression Fraud (Ad stacking, Ad Stuffing)

Ad stacking is another very common form of fraud. It occurs when
publishers stack ads on top of one another, and even though only the top
ad is shown to the user, advertisers think that their ads are also getting
shown. The image below shows an example of an ad blocker showing
how many ads are being served on the page or being attempted to being
served on the page.
They do this by placing ads in invisible iframes with zero to zero pixels and
zero visibility. This way they can load lots and lots of ads on the page.
Furthermore, publishers can use bots to load the pages so that the ads get
“viewed,” costing advertisers for each impression “viewed.”
This enables publishers to sell a huge amount of inventory on ad

exchanges, which are sold but never seen.
Fake Sites & Domains
This one is pretty basic and is exactly what it sounds like. These are fake
websites built just for advertising which offer no content that anyone is
interested in. They usually are integrated into a complex web of other sites
and are part of a larger network so that they don’t rouse suspicion. And
just like impression fraud, they can use bots to load and reload the page,
making it look like they have traffic and costing the advertisers money for
useless impressions.
Ad Viewability
This isn’t a type of fraud. Rather, I see it more as a challenge that is
currently being acknowledged by key players in the industry.
In the past few years, statistics about display advertising have been a hot
topic. Why? Because the figures are shocking. For example, did you know
that you’re more likely to survive a plane crash than click on a banner ad?
Or that you’re more likely to get into MIT or Harvard than click on a banner
ad? Eeek! For someone who lives and breathes the digital advertising
world, these are horrifying statistics. But what I think is even more
daunting is that while global display ad spend is expected to reach over
$70 billion in 2016, over 56% of all ads served are not actually viewable!
So what makes an ad viewable?
According to the IAB, at least 50% of an ad must appear in the user’s

browser window for at least 1 second, and in the case for videos, at least
2 seconds.18 That means if your ad shows up on someone’s browser, but
is placed below the fold, and the user hasn’t scrolled, there is no ad
viewability. To put it simply, just because an ad is served doesn’t mean
it’s viewed.
Although all of these challenges did exist in the pre-programmatic era,

they have been exacerbated by the current conditions of the industry,
specifically the decline in human to human contact. Fraudsters can easily
hide their identities and motives to remain undetected by brands and ad
exchange. In order for programmatic to overcome fraud, an audit trail
must be incorporated in the supply chain. There must be more
transparency in the system and a way to recognize if players
participating in the programmatic game should be there in the first place.
CHAPTER 9
FOR PROGRAMMATIC
I could not conclude this e-book without essential tips and tricks
as to how you can enhance your media buying strategies in
regards to programmatic buying. And the key to programmatic
buying lies in Marketing Intelligence.
What is Marketing Intelligence?

Marketing Intelligence is the process of gathering all the world’s online media data
and presenting it in a useful and accessible way to allow for better decision making.
Marketing Intelligence for Media Buyers

I am about to show you exactly how you can use Marketing Intelligence to
overcome your programmatic buying challenges as an Advertiser or Agency.
In the next few examples, I will be presenting several different brands from
different industries in order to demonstrate the broad range of advertising
information that you can extract from AdClarity, the essential tool for digital
advertisers and media buyers.
Challenge 1 Discover the Most Successful Sites to Advertise On
In this scenario, let us assume that we are a mid-size advertiser

in the travel industry. If I want to make sure I am targeting the right
publishers, I can do so by monitoring my competitors and top
influencers in the industry to find out what works best for them.
Let me show you how:

As an advertiser within the travel industry, I will use AdClarity to look up
TripAdvisor, my industry leader and uncover their online marketing
strategies and advertising tactics to create my own successful banner
ad campaigns.
In addition to seeing which publisher sites TripAdvisor is advertising

with, I can filter the sites according to "highest share of voice” to see
which publishers they are working the most with and therefore spending
most of their display ad budget with. I can also assume that the most
invested sites are deemed to bring in the best results because
TripAdvisor is not likely to be wasting such a huge amount of their
budget on sites that they know are not working for them.
The search results here bring up a long list of successful publisher

sites who are pursuing the same target audience as I am. Therefore I
can guarantee that there are several sites on this list which I haven’t
even thought about advertising with, and now I can expand my reach
to find new target audiences.
Challenge 2 Optimize Your Banner Creatives and Campaign Messaging
None of the above matters if you’re not able to speak to your

audience. And the way you speak to your audience is through
designing a creative that communicates, resonates, and
interacts well with your target audience. But which creative is
guaranteed to work?
In order to get an idea of what kind of creative design and

messaging you should be using, take a look at what is working (and
what is not working) for your competitors and industry leaders.
This time I am in the financial industry and I need to figure out what
kind of call-to-action, banner ads and landing page design my
creative needs to have in order to generate the highest conversion
rates.
All I need to do is uncover American Express’ online advertising

tactics to see which creatives are most successful for them and
build off of their campaigns to create my own ultimate campaigns.
Challenge 3 Finding the Most Relevant Mediators
Find the most relevant Ad Exchanges, Ad Networks, Ad Servers, DSPs,

and SSPs that will lead you back to your favorite publishers and other
similar publishers who target your ideal audience.
In this scenario, let us assume that we are a brand in the automotive

industry who wants to advertise on turbododge.com, an online publisher site
that deals with the automotive industry.
By using AdClarity to discover which mediators (ad exchanges, ad servers, ad

networks, SSPs, and DSPs) turbododge.com is working with, I am able to know
who I need to out to in order to advertise with them.
Additionally, I can assume that publishers in the same industry as

turbododge.com are most likely using the same SSPs. I should therefore make
sure that the ad exchange I am working with works with that specific SSP or
that I’m using the SSPs that are most relevant for me.
I can also use AdClarity to figure out which DSP I should be working with. By
using Marketing Intelligence to uncover which SSPs and ad exchanges a DSP
is working with, I can see which DSPs have access to premium inventory and
if other advertisers they are working with are in the same industry as me.
If I want to dig a little deeper into that last step, I can use AdClarity API
to evaluate the ad spend of my competitors and what they’re bidding for
inventory on a publisher site I'm interested in. In order to get the ad
spend, I calculate all the publishers they are working with, how many
impressions they bought from each publisher, and what the CPM is of
each of these publishers.
Marketing Intelligence for Media Sellers

In the next few examples, I will be using several different companies
from different industries in order to show you the broad range of
information that AdClarity has access to.
Challenge 1 Finding the Relevant Mediators
Find the relevant ad networks, DSPs, and SSPs that will lead you back
to your favorite advertisers and other relevant advertisers in order to
maximize your advertising revenue and access thousands of
advertisers that wouldn’t have bought directly from you.
In this scenario, we are a publisher who wants to work with Travelzoo.
By using AdClarity to discover which mediators (ad exchanges, ad servers, ad

networks, SSPs, and DSPs) Travelzoo is working with, I am able to know
which mediator I need to be in contact with in order to work with them.
Additionally, I can assume that advertisers in the same industry as

Travelzoo are most likely using the same DSPs. I should therefore make
sure that the ad exchange I am working with works with that specific DSP
or that I’m working with the DSPs that are most relevant for me. This will
help me find advertisers that will yield results from my site and will
continue running their ads with me.
I can also use AdClarity to figure out which SSP and ad exchange I should
be working with. By using Marketing Intelligence to uncover which DSPs
and ad exchanges an SSP is working with, I can see which SSPs have
access to premium advertisers and if other publishers they are working
with are in the same industry as me. By making sure I have access to top
of the line advertisers, I can maximize my revenue by knowing that premium

advertisers are trying to outbid each other for my premium inventory.
Additionally, AdClarity shows me the eCPM of my competitors so that I can

see what my inventory is really worth and whether my price floor is too high
or too low. If it’s too high, I’ll know to lower it so that I won’t be missing out on
advertisers. And if it’s too low, then I can raise it and maximize revenue.
In conclusion, Marketing Intelligence lets you gain complete visibility and

transparency onto your competitors advertising tactics and monitor your
entire ecosystem to build on their success and avoid their failures.
Your competitors have spent an exorbitant amount of their budget on

testing and trying to figure out what brings them the best results and
these insights will allow you to easily work off their successes.
CONCLUSION
CONCLUSION
Programmatic buying is indeed the renaissance of traditional media
buying. And like all good things in life, it shouldn’t be something that
is taken for granted. Programmatic brings a lot to the table in the
online advertising industry, yet its complexity and esoteric nature
requires all participants to really understand what it is and to stay up
to date with the latest innovations.
Each player discussed in this guide plays a key role in the

functionality of programmatic and none should be overlooked. As the
technology behind programmatic continues to progress, the
programmatic industry will keep evolving. And as this unceasing
evolution continues, it will eventually address and hopefully overcome
the challenges and obstacles facing the industry, providing an optimal
media buying and selling environment for all parties involved.
We hope you enjoyed your Ultimate Guide to the

Evolution of Programmatic Buying. If you have any
further questions or would like to learn more, don’t
hesitate to contact us at marketing@adclarity.com
CONCLUSION
Sources:
1. http://www.nlm.nih.gov/medlineplus/ency/anatomyvideos/000016.htm
2. These data providers are also applied to publishers, but for now we are focusing on advertisers
3. http://www.business2community.com/digital-marketing/retargeting-role-first-party-data-01132664
4. http://www.business2community.com/big-data/benefits-potential-pitfalls-third-party-data-0604309
5. http://blogs.wsj.com/cmo/2014/10/06/forrester-us-online-display-ad-spending-will-nearly-double-by-2019/
6. https://econsultancy.com/blog/66202-more-reasons-why-marketers-must-move-to-a-mobile-first-ad-strategy/
7. http://www.highervisibility.com/12-mobile-statistics-to-get-you-amped-for-2015/
8. http://www.emarketer.com/Article/US-Programmatic-Ad-Spend-Tops-10-Billion-This-Year-Double-by-2016/1011312
9. http://www.forbes.com/sites/niallmccarthy/2014/10/29/mobile-app-usage-by-the-numbers-infographic/
10. http://www.adweek.com/socialtimes/facebooks-q4-2014-mobile-is-now-69-of-ad-revenue/613826
11. http://adage.com/article/digital/facebook-s-mobile-revenue-hits-2-5-billion-prices-soar/296869/
12. http://finance.yahoo.com/news/why-google-losing-mobile-search-130007811.html
13. http://searchcrm.techtarget.com/feature/Location-based-apps-present-opportunities-and-data-challenges
14. http://www.adweek.com/news/technology/7-things-you-need-know-about-bots-are-threatening-ad-industry-161849
15. https://econsultancy.com/blog/63818-what-is-click-fraud-and-how-can-you-prevent-it/
16. https://www.exchangewire.com/blog/2014/05/29/the-truth-about-online-ad-fraud/
17. http://www.iab.net/media/file/IABDigitalSimplifiedUnderstandingOnlineTrafficFraud.pdf
18. http://www.iab.net/iablog/2014/03/viewability-has-arrived-what-you-need-to-know-to-see-through-this-sea-change.html
AdClarity
Brought to you by AdClarity, the leading
Marketing Intelligence tool in the industry.
REQUEST A DEMO

The Ultimate Guide To The Evolution of Programmatic Buying AdClarity E Book

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Ultimate Guide To The Evolution of Programmatic Buying AdClarity E Book

Uploaded by

Copyright:

Available Formats

THE ULTIMATE GUIDE TO

In this e-book, you’ll learn what

We hope you enjoy your read

As time went on and the number of advertisers and publishers grew

And history repeated itself.

And thus, the Ad Exchange was born.

Simply put, an ad exchange is a technology platform that facilitates in the

And although ad exchanges made things better, advertisers and

So along came Demand-Side Platforms (DSPs) and Supply-Side

results and data from DMPs. It could be seen as a sort of ad

So now you understand the key players in the programmatic buying

DSPS AND DMPS

In the previous chapter, I mentioned that the DSP acts

First Party Data

Second Party Data

Third Party Data

Example 2 Combining third party data with ﬁrst party data

Someone came to your site to buy a TV and you placed cookies on

Example 2 Using isolated third party data

You’re a sewing machine company who wants to get new customers.

Uhh… WTF is a regression analysis?

A regression analysis is simply a way of using one variable, or an input,

The three variables are:

Let me draw it out for you.

Ok, so how does an ad network actually work?

It sounds ﬁne…so what’s the problem?

instability, their packages were inconsistent and they were either

It eliminates the traditional methods that provoked wasted impressions.

The key differences between ad exchanges and ad networks

2 Advertisers can now buy only the impressions

2 Transparency for all! Because ad exchanges

4 If you remember, when I introduced the ad

PROGRAMMATIC BUYING IS THE AUTOMATED PROCESS

Good. Now that we have that settled, let’s begin.

So what the hell is RTB?

Real-time bidding is the buying and selling of ad space (per impression) in

Okay… so how does it work?

over to an ad exchange, the ad exchange could be working with several

Advertisers no longer have to work directly with publishers and negotiate.

But I guess that also depends on whether the glass is

So here’s the deal.

Before programmatic came around, advertisers cared about publishers. They

But all of this changed with programmatic.

A few things happened because of this.

How? Through retargeting.

So on the one hand, there is a huge debate on whether RTB is hurting or

programmatic guaranteed programmatic reserved

Programmatic Direct premium direct

And the publishers are fully aware of this.

As a premium publisher [I believe] the advantage goes to the

Buyers beneﬁt more because programmatic “allows them to

Buyers are beneﬁting more, as they are able to get massive

So we see that although RTB is a stellar technology, it still has some

And this is where programmatic direct comes in.

So what is programmatic direct?

Programmatic direct is the automated process of direct media buying

Another common notion is that programmatic direct is only about

Let me try to break it down in simpler terms.

player in the decision factor. However, programmatic direct provides the

In my opinion, programmatic direct is primarily for brands, not

So how does it work?

All programmatic direct is, is the automated process of traditional direct

The advertiser works with a publisher to buy a packaged deal of premium