Ad Targeting: Customer Data Collection In Advertising

Ads bombard users

Have you ever stumbled upon an ad that creeped you out as if the advertiser could read your thoughts?

Have you ever noticed that you’ve been getting same ads from several different sites on your PC, phone and tablet?

Ad targeting techniques increasingly rely on data collection and analysis by ad tech companies and marketers—all contributing to that uncomfortable feeling of being watched too closely.

In this two-part series, we’re raising the curtain on behavioral targeting technology, data-driven advertising, and data collection. Most users don’t know about these practices, not to mention the specifics of technologies powering them. Ironically, as Internet users, most of us and our purchasing decisions online are affected by data driven ad techniques, even though we may not notice this. This article will help to understand the mechanisms behind behavioral targeting, clarify the privacy risks and consequences of such methods to the end users.  

The Role Of Ad Targeting In Online Advertising

The benefits of online ad campaigns for marketers are evident: massive reach, a plentitude of ad formats, the possibility to measure multiple metrics for basically every step of the marketing funnel. From the first ad impression through to the conversion from prospective to customer on an advertiser’s website—all these steps and their quantitative indicators are thoroughly analyzed by marketing experts to optimize campaign performance.

The principal goal of ad campaign optimization is to reach the desired marketing objectives (i.e. sales, site visits, registrations) at the lowest price possible.

Businesses use ad targeting to narrow down their audience to those who are most likely to commit to a certain action at a certain time. For example, an ad for grill supplies might surface on your Facebook shortly after researching barbeque recipes, summer party planning, and grill care.

Not only are these ads more likely to convert customers, they ultimately cost less. A reduction in ad spend is the indirect result of targeting implementation since these ads are significantly more likely to yield buyers than untargeted advertising.

The websites that are placing these ads benefit also, because they can sell different parts of their inventory to the most interested advertisers at the best price.  

The Ad Targeting Types

There are many, sometimes scary-specific kinds of ad targeting available to marketers on most major platforms, like Facebook, Google, and Twitter etc.

Some of the broader targeting types are:

  • Geographical — targeting audience from specific country, state or city. Some advertising platforms offer even more granular geo-targeting, down to the city block. It is also known as “hyperlocal” and is often used for mobile ad campaigns and Facebook.   
  • Placement-based — targeting specific websites, pre-selected from the list of available ones on the ad network. This type of targeting is essential for advertisers in case an ad network doesn’t offer full fledged contextual targeting solution across all of its websites. For instance, if you’re an advertiser selling boats, you’d be more interested in targeting sites that are related to hunting/fishing, since they are frequented by your potential customers.     
  • Time-based — allows advertisers to determine time slot or day of the week to show their ads.
  • Device-based — enables advertisers to display  ads based on users’ device, OS version or browser version.
  • Demographic —  targeting method used by advertisers to display ads to certain audience segment,  based on parameters like age, gender, marital status, education, etc.

Contextual targeting  was a major breakthrough in the world of online advertising. It allowed the websites to display ads similar to the content of web pages, increasing their relevance for site visitors. It is also used in search engine advertising, showing ads based on users’ search queries.

The Rise of Ad Targeting and Evolution to Behavioral Targeting

The continued evolution of advertising technology from non-targeted banners back in 1994, to pop ups and further to video ads, tracked by a set of parameters induced changes in customers’ response to marketing efforts. Users learned to ignore the ads, developing so-called “banner blindness.”  Competition for customers’ attention grew fierce, and to stay ahead, marketers needed to get their message to an increasingly focused audience. And it couldn’t be just a general message; it had to be delivered at the right time with the right tone.

Furthermore, retaining existing customers required a deeper understanding of their audience and sophisticated forecasting to predict the needs of clients. Businesses achieved this by implementing a new approach to customer tracking and data analysis that is presently called predictive behavioral targeting.

What Exactly Is Behavioral Targeting?

Behavioral targeting is a method of targeting ads at users based on their browsing patterns, including sites they visit, their searches, purchases, etc.

Behavioral targeting uses data analysis of the audience to categorize it into individual segments. Expectations are that users belonging to the particular segment will be more likely to respond to ads based on the predicted interests. It is worth noting that contextual targeting and demographic information are used in combination with browsing habits analysis. However, since the interests of the audience are decisive in behavioral targeting, the ads served on the website aren’t required to be contextually relevant to web page’s content.   

If this is starting to sound invasive, you’re not alone. But how do advertisers get all this information?

How behavioral targeting works.
How behavioral targeting works. Source:

Tracking: How Advertisers Know How to Target Users?

The most common way to track behavior and determine browsing habits is through cookie files stored by the browser during the session. When your browser requests a page with some content, server will reply with the HTML elements for loading on the browser along with the cookie text file that contains a unique anonymous id. This is saved on your hard drive.

Cookie usage by websites
Cookie usage by websites

This is why if you  return to the same website the next day, you’ll find that some products you’ve clicked on previously will be displayed again in “Recently viewed” section or in the cart if you added them. For every page you’ve opened, the web server has sent you needed information and a cookie.

When you revisit the website your browser will send the request and the cookies it stored. With the help of cookies, sites can “recognize” visitors and “recall” actions they performed during their visit. This is the reason why you see the “Recently viewed” section. The website “recognized” you and personalized the home page for you.

Cookies placed on the website are also known as first-party cookies. Personally identifiable information isn’t stored within cookies, they only contain anonymous data about site usage. In the case of doubt, a good idea would be to check the Privacy Policy of the particular website. First-party cookies may be used for so called on-site behavioral targeting, which on the whole may not seem too invasive, right?

But there is more to it.

Most websites have ad placements/content placements coming from various ad networks or other web pages hosted on different domains than the one you’re visiting. When these ads or content are loaded on the website, your browser saves cookies from those different domains on the hard drive. These are known as third-party tracking cookies.

Even though these cookies are set by different domain, not by the one you’re at, they’ll be able to tell more about your interests and browsing habits to a larger number of sites you visit on the web later on. Just like first-party cookies, they contain your unique id’s and site usage info,set by third-party domains.

Only a domain that sets a cookie may read information stored in it, so it should be impossible for other domains you go to learn anything about you. Unless these sites have something in common, like using same ad network to monetize their ad spaces.

The best way to understand how third-party cookies work is to look an example:  I went to WSJ. com and looked up the cookies downloaded by my browser. The first cookie is 1st party, coming from itself. The rest are either from ad networks or so called Data Management Platforms (highlighted is the cookie set by Bluekai DMP). In the table, the domain where cookie originates is not This is exactly what a third-party tracking cookie looks like. Judging from the data in “Value” column, my browser has been assigned some unique id. We can’t know what exactly is stored within the cookie, besides the fact it’s unique.

Since the cookie is now stored on my PC, Bluekai will be able to identify my browser and collect my behavioral patterns along with other data across thousands of sites they place their code at. Every time I visit one, my browser will send cookies including the ones set by Bluekai on  and other sites.

(Not really, I just deleted them).

First-party cookies set by and third-party cookies from Bluekai DMP
First-party cookies set by and third-party cookies from Bluekai DMP

How Big Is the Data Collection/Behavioral Targeting  Niche?

Just one example from a recipe website showing 54 trackers loaded from its main page may give you an impression. An overwhelming majority of trackers are coming from the third-party sites and are intended for ad serving. Besides some of the well-known businesses, there are quite a few that most users don’t know at all.

This raises a lot of trust issues between customers and websites because, in many cases, users aren’t explicitly notified of these additional trackers. Even though these companies assure that data is properly anonymized and aggregated, some privacy advocates argue there is still room for improvement, by adding encryption and the possibility for users to opt out of data collection in the first place.    

Third party ad trackers on an average website
Third party ad trackers on an average website

How Is Data Collection Controlled?

Marketers claim that behavioral targeting is making ads more “personalized” and, in fact, improves user experience. Privacy concerns and the rise in popularity of ad blocker software in recent years may hint that users aren’t all that excited about being stalked by well-targeted, “personalized” ads.
Recently, I posted a question on Quora, asking: Have you ever felt that advertisers/marketers are watching you too closely after seeing an ad online?. It turns out that people who work in advertising themselves aren’t ​too satisfied with their user experience. Here’s a part of an answer by TV marketer:

Marketer on behavioral targeting
Marketer on behavioral targeting

But, isn’t this regulated? Well, sort of, but hardly enough

Behavioral targeting and the interest-based ad space is regulated by the Digital Advertising Alliance, whose founding associations include the Internet Advertising Board, and a host of respected businesses, including Adobe and Amazon.

Alliance members create and enforce ethical standards and practices in all niches of behavioral advertising, advocating transparency and customer control. They also work to ensure  accountability and enforcement of standards by participating companies. Their Adchoices program, started in 2010 was the first instrument that actually helped users to control how information about their interests  is used for targeting ads.  

Self regulation doesn’t always live up to expectations of governments, and they actively engage in investigations related to data collection practices. Recently French regulators fined Facebook for EUR 150,000 when it was discovered that social media giant used personal information to target ads and collected information about users on third-party sites without disclosing it.  Despite the insignificant amount in comparison to Facebook revenues, it is clear that regulators send an unequivocal message to the whole industry demanding transparency and respect for privacy.

How to Avoid Excessive Third-party Tracking?