The End of Third Party Cookies for Advertising

Third party cookies

In January of 2020 Google announced that Chrome was FINALLY going to be ending support for third party cookies. This is a huge win for consumers and their privacy, and a huge complication for many advertisers who rely on advertising to specific users using that data.

Let’s start by defining cookies. A digital cookie is a piece of code (or script or small piece of software) that is embedded in your internet browser and collects some type of data and feeds it back to the owner of that cookie.

Cookies can do some great things that make browsing the internet a fast and enjoyable experience. The ones that do that are typically first party cookies. Those are cookies that are placed by a publisher (owner of a website you are visiting) and collect data about what you are doing on that specific website. That first-party data is then used to personalize your experience on that website. That can come in the form of saved carts for e-commerce websites, saved passwords, or remembering the links you previously clicked on. They can even be used to personalize your experience in the form of remembering your zip code and delivering weather that is most relevant to you.

First and Third Party cookies tend to get lumped together as all bad-guys, but if first party cookies went away tomorrow, we would all be clamoring for them to return.

Third party cookies are where things get more suspect. These are cookies that are placed on a website, but used to track your behavior across websites. The data becomes "third-party" when you are on one website, but a third-party is tracking your data on the website you are currently on.

An example would be if you navigated to a website like CNN. CNN may place a tracking cookie on your browser that does collect information about what you are doing on CNN, but it may ALSO follow you after you leave CNN and go to ESPN. Now that CNN cookie is a third party that is intercepting data about your browsing on ESPN (and anywhere else you go on the internet) to collect information that can be packaged as a "behavior" and sold to advertisers.

What is even more troubling is the type of data that some of these cookies are collecting. You might expect that they are collecting data about the types of websites you visit, how long you are on those sites, and the specific content you are browsing. But that cookie is probably also collecting information about your IP address, then being combined with off-line data from public sources to get more specific information about your household income, house value, vehicle registrations and more. They may also be able to track the information that you are entering into online forms to get even more information about you.

Seems crazy, right? The reason this type of technology has survived this long is the defense that the data is completely anonymized. This is true. Any PII (Personally Identifiable Information) is stripped out of the cookie and all the publisher of the cookie is left with is all your attributes. Every minute detail. Just not a way to draw your data back to your specific name or address.

What do advertisers do with third party cookie data?

As previously mentioned, advertisers can attempt to use the data to place users into "buckets". The buckets can range from general (sports enthusiasts) to specific (Red Model T enthusiasts). What’s more, advertisers can combine buckets to get even more narrow targeting (Sports enthusiasts WHO ARE ALSO Red Model T enthusiasts).

That is a dream for many marketers. The ability to define an audience that is best for your product or service and then show ads directly to those people. This ability to target is what has driven an explosion of advertising money into digital ad exchanges. (Side note: There still is a ton of question about how accurate this targeting ACTUALLY is, and most data companies who collect and provide these audiences are black boxes who hide behind "proprietary models" to protect exposing this accuracy/inaccuracy, but I digress).

What’s next?

If reading about the third-party cookie makes you feel as dirty as it made me feel to write it, then you can understand why there has been a movement in the industry to end its use. All of this came to a head in 2019 when Google said they were going to end support for the third-party cookie in Chrome. But they were hardly the first browser owner to end the support of the third-party cookie. In fact, Mozilla and Apple (Firefox and Safari, respectively) were way ahead of Google. Google made the big splash because of Chrome’s dominance in the marketplace, owning over 60% penetration of the browser market.

It is still up in the air about when exactly Google will stop the third-party cookie support. Most likely early 2022. A lot likely depends on the launch of their "Privacy Sandbox", which will be a tool kit that will supposedly protect consumer data, but still allow advertisers to do some somewhat specific targeting (albeit likely less specific than previously).

There are also some questions about how the end of the third-party cookie will impact attribution models. An example of an attribution would be: User gets exposed to my ad, then makes a future purchase on my website. I would hope to be able to attribute that ad exposure to the sale that takes place on my website. Tracking cookies make that happen today.

Bottom Line

As a marketer, especially one that is running digital display campaigns, you should be aware of how your ads are being targeted. If you are relying on specific data that you don’t own to target users, then that may get more complicated.

At Digiarks, we are huge proponents of businesses collecting and harnessing their own data. If you own data that includes physical addresses, you can entirely remove the third-party tracking cookies from your advertising equation by investing in IP targeting. With IP targeting you are targeting households, rather than specific users. While you lose the ability to target a specific user, the technology that targets the IP address is more accurate than the algorithm that places users into behavioral buckets anyway.

Or you can use your data on a 1-to-1 basis to target specific users on Facebook or LinkedIn. This is often called "customer match", where you are matching your list (based on a variety of factors like name, phone number, email address, and physical address) to specific users on these platforms.

Another thing to watch is how robust Google’s privacy sandbox will be. As the largest owner of first-party data, they will likely become even more dominant and relied upon for display advertising moving forward.

In the end, if you have questions about this complicated landscape, reach out to us! We are always happy to geek this out with you.

Go back