How GenAI will disrupt the multi billion dollar digital advertising industry

First let us learn how the digital advertising and online ads targeting model works. We are accustomed to seeing ads online that seem to intuitively follow what we last purchased or browsed. These ads are behaviorally targeted at us. I will briefly explain the key concepts in an online advertising ecosystem and how data flows among them.

Publisher: Refers to a website a user may visit and that can display ads.

Supply-side platform (SSP): Refers to the ad network that ultimately decided which ad needs to be placed for this user on the website they are visiting

Demand-side platform (DSP): Refers to the companies that work with SSPs to complete the fulfillment, so in our case display the ad to the user on the website.

Real-time bidding is a process that enables publishers and advertisers to deliver targeted ads online. In digital advertising, the basic unit is the "impression," which occurs each time someone loads a web page displaying an ad. When this happens, the advertiser pays the ad network to display their ad, and in turn, the ad network compensates the publisher of the website where the ad appears. Before an ad is displayed, a rapid auction determines which ad will be shown. During this split-second auction, a user’s information is shared with multiple bidders, and automated systems on various corporate servers compete in real time to win the chance to present their ad to that user. Personal data moves from the user to the bidders, while payment for the winning impression flows back to the publisher.

Picture credit : https://www.eff.org/wp/behind-the-one-way-mirror#Part3

To summarize, as I keep scouting for a new face wash with the least harmful ingredients at the most effective price point, I am targeted with more and more skin care products on every browser tab I open, on my social media feed and on every website I visit. Tracking technologies such as cookies, pixels, web beacons assist with online analytics.

Now enter the new world of GenAI. Perplexity is my favorite go to at the moment. A few weeks ago I searched this prompt on Perplexity -

Compare the ingredients between Estee Lauder Soft Clean Moisture Rich Foaming Cleanser, Chanticaille rice geranium foaming cleanser, 9wishes Rice Foaming Cleanser, Augustinus Bader The Foaming Cleanser.

Perplexity presented to me a neat tabular chart comparing the ingredients and then it gave me this option of buying with a pro account. All I need to do was connect my credit card. Suddenly I don’t have the need to go to any another website to place my order, which means the DSP platform lost an opportunity to serve me a skin care product ad.

This is also scary that my AI assistant has gradually become very powerful and invincible by learning my likes and dislikes. Now my goto AI assistant or my AI agent has built a rich customer profile of my skin care products preferences based on my search and my behavior. This has been the holy grail of the marketing industry to build a 360 degree customer profile to target the right customer with the right offer at the right time. GenAI and AI agents can easily solve this by enriching the customer profiles through personalization.

Traditional tracking methods like cookies and pixels may decline in importance as GenAI leverages AI-driven insights and modeling to replace direct user-level tracking with broader predictive analytics and contextual understanding.

At the core of this will be the policy of these big companies that own the AI on how they should or should not use the customer’s data also known as purpose of processing. The saving grace will be the ethics behind that policy.

I conclude by citing this analogy, a thief, a locksmith and a policeman all are well-versed at picking locks. They do it for different reasons.

Here is a little youtube short I made for this post.

Now does it make more sense as to why someone would offer $34.5 billion to buy Google's Chrome web browser? 😉