Mash Snack: 3rd Party Cookies

Season 2 episode 6 | March 11, 2021

Mash Snack: 3rd Party Cookies

March 11, 2021


Now that 3rd Party cookies are going extinct on the Chrome browser, what new fun thing will replace them?

Audio Transcript

Last year Chrome announced its intent to remove support for third-party cookies, as a way to pivot their privacy efforts in favour of their users. The reality is, that this means Google has already found another way to use non identifiable data to use for advertising targeting. When this first came to light, we all assumed that Google Chrome’s privacy changes could impact some areas of the marketing and advertising space. How to track conversions for one – which is a huge part of the 3rd party cookie’s usage for people like us, at Mash.
Generally, we harness the power of this data (we don’t see it or store it, it’s all kept secret from us as well) to pinpoint our clients audiences, and use the machine learning aspect of the data’s use in our marketing strategies in paid ads. So how do we navigate these changes?

Google’s new privacy code is called The Sandbox – From google “ Sandbox will build innovations that protect anonymity while still delivering results for advertisers and publishers. once third-party cookies are phased out, we will not build alternate identifiers to track individuals as they browse across the web, nor will we use them in our products.” – this is a quote from David Temkin at Google, who is their director of product management.

The proposed changes include something called Browser fingerprinting, which is the practice of gathering enough information about a specific browser instance to try to identify a user in a cohort, instead of a unique user. They call it a FLOC or Federated Learning of Cohorts and in the case of say, Google Ads generating interest based audiences to target, it’s an effective way to measure it, well at least 95% of conversions per dollar spent compared to 3rd party cookies.

How does that work for conversion tracking? Well one of their new APIs is something called an event level iteration, and they recommend using Google Tag manager in the meantime as that will mean a smoother transition. Unfortunately there is no detail as yet how the new product or procedure will be built, we will have to wait for a testing prototype. I believe if you are signed up as a beta tester for Google’s APIs you may get an invitation to test this sooner than consumer level, but check back with the Google Blog about access there.

Now, this technology isn’t entirely new, well, the IDEA of it isn’t new. Remember other browsers got rid of 3rd party cookies years ago, almost ten years ago in fact. 2013. Browsers like Safari used something called Campaign ID, which was just 6 bits of information to track when ‘someone’ followed an ad, not the 64 bit that Google currently utilises, and the impression data from Apple was unreliable, but when comparing.. well, Apples with Apples, you have a consistency at least.

So back to the FLOC – which as I mentioned was a ‘cohort’ based fingerprint. So if you aren’t sure, a cohort is a group of users in this case. And the idea is that Google will track your browser patterns wholistically, and assign you to a Floc cohort.

This means it’s kind of exciting – you know how in Facebook for example you can build an audience based on interest, demographics, location and somewhat on behaviour? Well this means for Google, it’s all wrapped in one. Your behavior on the internet might not identify you as a person, but your FLOC cohort will pigeonhole you based on what you read, search for, buy, etc. You’ll be grouped in with people who almost exactly behave the way you do online.

I say exciting as I can totally see how this will assist with interest based targeting in a Google Ads environment and how it granulates the way we can serve ads.

My big question is, can these Cohorts be used for something evil? For example – We see discrimination in our real life day to day dealings with businesses and individuals, so can belonging to a specific cohort of users be detrimental to what you are shown online? Can you be excluded or persuaded based on your FLOC? These will no doubt be used in their machine learning and one example I can think of is a cohort who searches payday loans or financial hardship can be excluded from some advertising of products or services when they might actually be relevant customers. It could get nasty when peer to peer lenders use this data to push high acceptance, high interest loans on them.

Anyway, there is a lot to unpack on this one, and not all the information has come out yet, especially for us marketing people – if you’re interested in playing with the Sandbox and reading more about the APIs, there are articles and access information on the Google Blog, I’ll drop the links below:



purple line

Discover our latest Podcast episodes