“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Ari Paparo, CEO at Beeswax.
"You finally really did it. You maniacs! You blew it up!" – Charlton Heston, “Planet of the Apes”
Personally, I was a skeptic on whether or when Google Chrome would finally kill the third-party cookie. But that day is here, or, I guess, two years from now it will be, but you get the point. Let's go around the bases and see what this means.
The new world of digital
Google announced that third-party cookies would be blocked in Chrome within two years, at which point they will be effectively dead for all use cases. Google also has been clamping down on fingerprinting techniques, which means there's no secret workaround. We've been operating our digital advertising businesses with a roughly 30% blind spot due to Safari, and that's now going to 100%. It is somewhat remarkable that the 30% threshold was not enough to cause major operational changes among buyers and sellers, but now we're confronting the reality of a world with significantly less measurement and targeting.
Google's plan moving forward is hinged on various concepts outlined in its Chrome Sandbox. In simple terms, these proposals use a browser sandbox, not accessible to the advertiser, to store signals such as clicks or conversions and strictly measure out those signals in ways that anonymize the user. The direct signal from the user's browser to the ad tech/marketing cloud is severed, and instead the browser does some measure of limitation, obfuscation and anonymization before any data is returned.
For example, in this draft proposal for click conversion, clicks are registered from ads and associated with an advertiser, and conversion events are fired from the browser, but in both cases without cookies. The conversion attribution is then done by the browser's sandbox and reported back to the advertiser in a batch, such that the individual user is difficult or impossible to isolate, with limited metadata about the source of the clicks.
Let's start with the deadpool:
- View-through attribution: dead.
- Third-party data: dead.
- DMPs: dead.
- Multitouch attribution: dead.
The removal of view-through attribution in particular is likely to cause the biggest change to media buying habits as it is a mainstay of most digital display measurement. It is hard for me to say this, but it seems like we will be moving back to last-click as the gold standard for attribution (sorry, I vomited on my keyboard). Although, it is worth noting that virtually all in-app attribution is last-click, and they are sustaining a multibillion-dollar ecosystem.
Some techniques that are currently cookie-based are less likely to die entirely. Measurement of reach and frequency and frequency capping are both likely to survive in some form, using either publisher first-party data or IP addresses, or some other approximating technique.
Targeting and optimization of digital campaigns is likely to get simpler, as inventory attributes (site, placement) become the clearest signals that can be correlated to the limited conversion and click data.
With no ability to measure view-through, the media mix will likely bifurcate to click-based direct response ads on the one hand, and video-heavy branding ads (with limited KPIs) on the other. There's a lot less room for hybrid "brand direct" spending when neither audience nor results can be clearly measured.
When I worked at Google there was a fatalistic engineering joke: If you're choosing a database to build on, you have a choice between one that's not ready yet, and one that's hopelessly outdated.
Google did no one a favor by putting a two-year horizon on cookie removal. The interim period will effectively freeze many businesses that rely on cookies in an uncertain state where they can't really assure customers and investors that they will be able to continue operating (since the solutions aren't ready yet) and they can't go on with business as usual (since the underlying assumptions will all change).
For publishers and advertisers, there will be a willingness to adopt some of these privacy sandbox features when they are fully vetted, but the transition will require heavy lifting as publisher and conversion tags will change, content management systems may need upgrades and reporting systems will certainly be radically different. Maybe some smart entrepreneurs are already working on the "Sandbox compliance-as-a-service" businesses.
What's not changing
We still don't know what the future holds for Apple's IDFA or Android IDs. We can surmise they are headed in a similar direction, but maybe over a longer time frame.
We also know that publisher first-party data is growing in importance, and that efforts to use this data in collaboration with advertisers seems like a promising area of investment. Whether this takes the form of hashed email addresses, shared logins or other technologies is still an open question.
And, of course, we still have …. cue the angels singing … connected TV.
A glimmer of hope
The most interesting part of Google's announcement is the idea that the sandbox techniques for ad tracking could become standards instituted in multiple browsers. Presumably, this would include Apple, the digital wasteland.
[Carrie Bradshaw's voice] Maybe a future where there is less tracking, but it is consistent across all browsers and even all apps, is a better world for ad tech than the balkanized state we currently find ourselves in.
But I wish we didn't have to wait two years to find out.