Forget cookieless. Now it’s time for ID-less solutions. No doubt you’ve heard someone drop that buzzword in a conversation recently.
But what is an ID-less solution, exactly?
As with many ad tech terms, the lack of a standard definition for ID-less solutions has created general confusion, said Shailley Singh, executive VP of product and chief operating officer at the IAB Tech Lab.
And so the Tech Lab did what the Tech Lab does.
It convened members of a working group – in this case, the one focused on addressability and privacy-enhancing technologies – to develop guidance that clearly defines what an ID-less solution is, outlines the main use cases and sets out a framework for how to evaluate these tools.
The Tech Lab released its draft guidance on ID-less solutions on Thursday, along with a call for feedback. The public comment period is open until December 19.
The hope is to get this guidance finalized before the end of the year, but if the Tech Lab gets a large volume of comments, Singh said, it probably won’t finish up until early next year.
When less is more (privacy-enhancing)
Demand for ID-less solutions is on the rise for obvious reasons: ongoing identity signal loss, regulatory scrutiny, privacy-focused browser changes and a desire to maintain some form of addressability.
“We’re seeing a lot of momentum,” Singh said. “But it’s a portfolio of different types of solutions, and each company will have to test these methods themselves.”
And before companies can start testing, they need to understand what they’re testing.
“ID-less” might sound self-explanatory – i.e., various methods for targeting ads and measuring campaign performance without individual user information – but there’s important nuance there.
For example, one might assume probabilistic identifiers are ID-less, because they’re created using a set of inferences that alone can’t identify a person. But probabilistic IDs can be associated with a specific user, and so the Tech Lab doesn’t consider products that use them to be ID-less.
Pros and cons
An ID-less solution, according to the Tech Lab, is one that preserves the anonymity of an individual while also supporting addressability.
Cohort-based targeting, like the APIs in the Chrome Privacy Sandbox, and good old contextual targeting are the two main types of ID-less solutions. But lots of other stuff falls under the ID-less umbrella, including the use of discount codes, aggregated attribution reporting, media mix modeling (MMM), brand lift studies, Seller Defined Audiences, on-device auctions and on-device frequency capping.
They each have their pros and cons.
Seller Defined Audiences, for example, which is a Tech Lab spec that publishers can use to monetize their first-party data on the open web without a user ID, supports standardized and custom audience segments and is cheaper to transact on because the data is available directly within bid requests.
But, as the Tech Lab acknowledges, bad actors can scrape first-party user data from RTB requests to build profiles, while SSP traffic-shaping rules may actually discard the data. (This is why we can’t have nice things, etc.)
MMM, meanwhile, which has come back into vogue as a result of signal loss, is more privacy-protective than other forms of measurement since it doesn’t incorporate user-level data and has a limited risk of abuse. Marketers can use MMM to attribute across any channel, including out-of-home. But it’s also less precise, takes more time and can be quite expensive.
And then consider on-device auctions, which enable more reliable retargeting than other tracking methods and protects against reidentification. However, it’s also limited because interest groups can’t be combined. Also, cross-device tracking isn’t supported, and the methodology requires much greater publisher adoption (looking at you, PAAPI), or it might not result in successful bids.
The only way for companies to know if an ID-less solution will work for them is to get busy testing it – and there’s no shortcut, Singh said.
“This is new technology and a new paradigm,” he said, “so you can’t really get around not testing or not doing a pilot.”
Change ain’t easy
The ID-less solutions themselves are looking forward to getting their tires kicked – way more so, perhaps, than some in the ad tech industry are excited about kicking their own long-held ID-based habit.
“Many industry leaders – brands, advertisers and agencies – hesitate to adopt ID-less solutions due to inertia, misconceptions and a focus on short-term goals,” said Abhishek Sen, CEO of NumberEight, a mobile data company that uses on-device computing to create cohorts that don’t rely on user-level identifiers.
NumberEight’s CTO Chris Watts was one of the main contributors to the Tech Lab’s ID-less solutions draft guidance through the Addressability and PETs Working Group, along with representatives from Anonymised, ThinkMedium, TripleLift, ID Privacy, Covatic, The Trade Desk, as well as Dstillery’s former principal engineer Brian May.
Change is hard, especially when it challenges a foundational and well-established practice, but “the shift is already underway,” said Sen, pointing to the Chrome Privacy Sandbox and Apple’s SKAdNetwork.
“The shift isn’t easy,” he said, “but those who act now will lead the industry into the future.”