When Google announced another delay in the phase-out of third-party cookies, the company said it was due to feedback from the ad industry, which said it needed more time to test and evaluate the alternatives being developed in Privacy Sandbox.
Some media professionals believe Google’s delay is a good thing, as it gives the industry more time to get the transition right. Walter Flaat, Dentsu Canada’s chief data officer, is one of them.
Flaat says he wouldn’t be surprised if Google’s timing shifts even more, since some of the tests of alternative options might not fully succeed, leading to a standstill period while regulators evaluate the solutions. “I’m almost looking at this as building a cathedral. The timeline shifts and you want to do it right the first time for the building not to come toppling down.”
He also expects to see privacy continue to be used by some as a business model – not just trying to be compliant, but to use strict privacy as a unique selling point. He says we’ve already seen this with the likes Apple, which has made privacy a key brand attribute and its iOS a “pay a premium for privacy” platform.
In a cookieless environment, Flaat says there are still going to be a few ways to work with third-party data. One is by buying into or moving onto a “digital continent” like Google, Meta or Amazon (which all hate the term “walled garden”).
“Within those environments, you’ll have some amazing capabilities to use data in sterile clean rooms for analytics, and you’ll be able to use these companies’ data for media activation,” Flaat says. “Those are all table stakes. All that data – and much of those platforms – will be effectively available to everyone. You need to be there to succeed but ultimately, it’s also very much a level playing field.”
For at least the last 18 months, Flaat says he has seen leading companies set themselves up for privileged access to third-party data in two main ways: partnering up to attempt sharing data, in clean rooms, with other enterprises; and non-competing partners gathering co-opt-ins on activations, so they both use data gathered from this as first-party data.
He says another interesting and valuable route is to look at some third-party, or even non-personal data a company may already have. Very often modeling techniques can be used to enrich first party-data, adding psychographics, media imperatives and many other data elements, which effectively become first-party data.
Less effective strategies for third-party, says Flaat, is trying to undo or work around the move to cookieless.
“That doesn’t make sense. Cookies are going away because they technically simply make absolutely no sense – it’s a messy hack that survived for decades and they do things that consumers and governments won’t accept. Instead of trying to redo a tracking pixel, start developing new, competitive audience strategies. Lean into better creatives and get ready for the new mechanics of media like RTB auctions not happening on an exchange, but on someone’s phone.”
Looking at the best method to target audiences in a cookieless future, Flaat things there will be a best-to-worst cascaded approach. The best will be based on first-party data customer matching that are one-to one matches with third-party data based on stable identifiers. Where that’s not possible, look at digital signals that still can be captured, such as people visiting a client’s website.
“When you absolutely don’t know a person – they’ve never been around on your site or app – you’d resort to targeting on groups of people,” Flaat says. “This could be based on users interested in a certain topic. There will be new APIs available that allow you to do this. But also: contextual. The age-old wisdom of serving ads related to the content they are shown around.”
