Programmatic Marketing and Role of Data Analytics and Artificial Intelligence

Evolution of Marketing
Marketing is the act of informing a prospective consumer about the
availability/features of something he is looking for and is ready to own by
paying a price.
With the passage of time and due to increasing competition, Marketing
Strategy has evolved from customers who are actually buying —> to
prospective customers who are looking for information to buy, —> to incite
the need/interest of buying in an ordinary person so as to convert him to a
prospective customer.
And, with the above mentioned evolving requirements and with the evolution
of technology, medium of Marketing has evolved from word of mouth —> to
print media —> to electronic media —> to social media and now in this
digital age —> to Programmatic Marketing.
In a race to gain more and more customers, Organisations have literally
started running after almost every person to gain attention, they want to be
present at almost every place and want to target almost every person. To
achieve this goal, the mediums of the widest reach are targeted. And even a
common person will agree that in today’s world at any point of time at any day
in almost any part of world, one pattern is common:
Number of people browsing internet >> Number of People Watching
Television >> Number of people reading Print Media.
So, where is the mass, where shall I advertise that maximum number of people
see my advertisement?
Obviously, it’s Internet. To advertise on Internet, a piece of code/program is
needed which can show and tell about the product being marketed. These
programs were written and this whole phenomenon of marketing on internet
gave birth to the concept of Programmatic Marketing.
So, Programmatic Marketing is advertising on Internet.
But the story doesn’t end here my friend, it actually starts from here.

Programmatic Marketing
If at any point of time, number of people browsing internet is maximum, does
it really mean that my advertisement on internet is being seen by all of them?
Answer is “NO”. Because Programmatic Marketing is designed to show
advertisement on websites, and are these people accessing 1 website or
different websites? Obviously, they are accessing different websites, and hence
if I want all of them to see my advertisement, I will have to show my
advertisement on all the websites.
Challenges faced by Organisations and the complexities of Programmatic Marketing
Q-1 Do we know what websites these people are browsing?
Q-2 Even if I know, will they again browse the same websites?
Answer of both the above questions is “NO”; this is so dynamic that it is
impossible to predict what websites consumers will browse.
Then, what is the solution. There are 2 options:
1. List all the websites hosted till date and advertise on them.
2. List the most browsed websites and advertise on them.
Option-1 is not possible because there are billions of websites most of them
being accessed very less and advertising on all of them could be a financial
disaster with negligibly low ROI.
Option-2 Yes, one can get the data of most accessed websites and advertise on
them. And, that’s what is happening in Programmatic marketing.
But the problem is not so simple. Even the most accessed websites are
humongous in number and are accessed for a purpose.
Do we just want to show our advertisement to each and every customer
accessing those websites? If yes, will it generate interest in the product or will
it be over advertising and kill the interest in it?
Definitely, showing every time will be overkill. Problem becomes even more
complex when the products to be advertised by an organisation are more in
number. Think of ecommerce websites, which product they should
advertise and which product they should not advertise?
So, what to do?

Answer is again programmatic marketing, it’s that program that will decide
which advertisement to show and when. For this, first the program needs to
know the purpose of the person accessing the website and then accordingly
decide the advertisement.
But how the program will know what is the intent of the person using that
Some will argue that by setting cookies and then tracking what the person was
doing. But is that really true?
Let’s take a hypothetical example: a Person “X” searches for Shoes on a
popular ecommerce website and cookies are tracking it. After 15 minutes, he
opens another website let’s say: Then, can we infer that the
person is accessing because he is looking for shoes? In
reality the answer is “NO” but in the world of Programmatic Advertising,
answer is “YES”. And, in this specific case, this person “X” will be shown Shoes
advertisement on the other website, thinking that the person might come back
and look for shoes again.
Will this person “X” will click on this shoes advertisement on and go back searching for shoes? Most probably not and
reason is that he might be done with searching of shoes or more appropriately,
his intent of opening was not for looking shoes.
So, is it a wrongly placed advertisement? Well it’s debatable.
Is it possible to know the intent of the person accessing the website each and
every time with 100% accuracy?
In some cases it is yes, for example on ecommerce website if a person is
searching for a laptop, he is actually looking for a laptop. And, if at that point
of time he is shown advertisements related to laptop, there is a very-very high
chance that he will click on that advertisement.
But, in most cases it is no. So, what is the solution?
There are 2 solutions:
1. Purpose of that website itself. In our earlier example, person “X” opened, then advertisements related to financial instruments
or advertisement of a bank offering higher rate of interest on FDs is
more apt then the shoes advertisement.
2. Data Analytics.

Role of Data Analytics in Programmatic Advertising
Data Analytics, How?
What if there is a pattern of accessing websites? For ex:
1. In early winters people start looking for heaters, woollen wears or
stuff related to winters.
2. In March or at the time of bonus payments, People start looking for
new cars?
3. In December, please start looking for holiday deals?
These are obvious but there will be certain hidden patterns in the way a
typical person accesses internet at a given time, day, month of a year. If we
analyse our usage pattern, our buying pattern, most of us will be very
consistent in the way we browse internet, the way be buy things etc.
Same thing is applicable on regions also. For example, on a given day, at a
given time, in metropolitan cities, there is a high chance that users might be
searching for high end electronics or services like urban clap etc. While in
villages this will be entirely different
So, Data Analytics based on this will reveal great amount of valuable insights
on at what time, in which region, what is the need to consumers. Various
debit/credit card issuing companies, search engines maintain a great deal of
this data, use this to generate these kinds of insights for Programmatic
Role of Artificial Intelligence in Programmatic Advertising
Let’s take a hypothetical example. Think about a domestic Airport Terminal
having multiple departure gates. Near each of the departure gates there is a
big LED TV Screen specifically for advertisements.
Typically, it will show a set of advertisements getting repeated again and
again cyclically.
Typically, near that departure gate, people departing to the city to which that
flight is going will stand and will see that LED TV.
But, what if the LED TV Screen is intelligent and knows that from its departure
gate flight to which city is departing? (Technically, this is extremely easy and
possible) or in other words where are the people watching it are heading up?
If that LED TV knows that it can use that information to show the
advertisements related to that city. For example: Hotels near that city’s

airport, tourist services in that city, best restaurants, clubs in that city or
anything relevant to the context.
Then, in which scenario the advertisement will have more impact:
a) When LED TV was showing cyclically repeated advertisement?
b) When LED TV showed advertisements based on context?
Obviously, it’s when it showed based on context.
There can be a great number of scenarios where Artificial Intelligence can be
used to do high impact programmatic advertising.
And now, if the data analytics suggest some high spend by volume and by value
in those cities and advertisements of those cities are further selected based on
that analysis, one can imagine what will be the impact of advertising, what
could be the hit rate.
Summing Up
Programmatic Marketing is a great tool to reach maximum number of
consumers but the biggest challenge is, if it is used as it is, it will be dumb and
direction less resulting in less ROI.
But together with Data Analytics and Artificial Intelligence it can do wonders.

Written by Ankit Gupta.

Ankit is a student of ePGP at IIM Ahmedabad.

Leave a Reply

Your email address will not be published. Required fields are marked *