Hi Darth,
Thanks greatly for the detailed analysis of the Unified ID and how it is different. And the CLEARCODE site does a good job of explaining things in more detail.
It was critical to know that cookies often only contain a unique ID and that the rest of the data is stored on servers at the advertisers and website companies. These IDs are what is mapped between advertisers and websites to refer to the same person, and allow them to share data they have each collected (Cookie Syncing).
The Trade Desk’s Unified ID solves the problem of matching each DSP’s ID they’ve assigned to a user’s cookies, a problem that currently causes lots of mismatches and requires each DSP to keep a map between their IDs and every other DSP’s IDs.
https://martechtoday.com/why-the-trade-desks-unified-id-may-…
In addition to being a massive matching effort, this process is highly inaccurate. Match rates above 60 percent are considered decent, which means that it’s common for a large percentage of user IDs to remain un-synched and therefore unrelated to other cookie-based history.
The worst privacy issues come from the use of Tracking Pixels. In simplified form, advertisers use Tracking Pixels to know when each ad they have served to a particular user (via website or email) is accessed by that user, and websites use them to track what a user is doing during a session on its website. Using Tracking Pixels in this way provides information about users without their consent, including the unique IP address of their computer allowing companies to identify the user activity whenever they access the internet. Spammers use this as well.
(Details on Tracking Pixels: https://en.ryte.com/wiki/Tracking_Pixel )
If it is Tracking Pixels that Google ends up blocking (already available for gmail as a Chrome extension called PixelBlock), then privacy is maintained and TTD’s Unified ID will not be affected. Things like PrivacyBadger (that enforces Do Not Track) appear to do that, but I haven’t been able to confirm that.
https://www.eff.org/issues/do-not-track
Darthtaco:So what the DSP uses to determine what ad to display (programmatic) is entirely and purposefully from a randomly generated and assigned ID.
I found a reference that confirms that:
https://martechtoday.com/2019-will-be-the-year-the-ad-tech-e…
One important point is that no actual data is exchanged in this process. The IDs that we pass and link together are completely random numbers, meaningless in a vacuum. All companies involved in the chain engage simply to exchange currencies – “I got ID123 on this user, what did you get? ID456? Great.” The seller doesn’t tell the DSP that the user is a soccer fan who streams a lot of games online. The DSP doesn’t tell anyone that the user is shopping for a new car. Both of those data points remain proprietary to their respective owners.
Darthtaco:TTD and others have created a system where big third party cookie reading has become less important. In place they have created a system where better targeting can be completed using only data in their own systems on a completely anonymized ID.
From what I have noted above, the websites still use a user cookie, but advertisers probably don’t.
Darthtaco:I hope my other post made that clear why that is true and why if Google stops transacting in identifiable consumer data it won’t effect TTD.
You sound very certain that it won’t affect TDD, and that is reassuring. But without knowing what Google is thinking of doing, I am concerned enough that I’d like to know how you are so certain.
Enjoy,
Brian