Facebook and Google are completely rebuilding how ad targeting works. What if they fail?
Will cohorts and other proxies have a chance against having someone's credit card data?
We’ve still got a month to go in Q1, and already we likely have the dumbest story in media and advertising in 2022, which of course is Applebee’s somehow getting in trouble for running ads on a news network. Congrats to all.
I was out of town last week, so take it with a grain of salt but somehow Applegate seemed to be getting more attention than the only story that really seems to matter- does digital advertising actually work anymore?
Or put another way - just how FUBAR is Facebook’s business?
I’m being a little facetious about the whole digital ads thing. They work - at least some of them still seem to. But I’m still trying to wrap my head around the still unfolding identity changes the business is enduring and the massive ramifications that have yet to full play out. As my old pal Brian Morrissey put it bluntly, it’s the End of an era of ad targeting.
That $232 billion one day loss in value. Facebook’s alarming guidance. And the absolutely killer quotes in this recent Wall Street Journal story that would be unthinkable a few years ago. It’s all still so stunning.
I find myself wondering, when you have 10 million advertisers relying on a targeting tactic that no longer works, the company acknowledges that it needs a plan B, and that the plan isn’t ready yet - what happens to what is essentially a micro economy?
“Facebook is basically starting over,” said Ari Lewine, co-founder and chief strategy officer at TripleLift. “Literally they will need to rebuild their ad stack some what from the ground up. Their existing data surveillance mechanism that feeds into their machine leaning algorithm is no longer going to be viable.”
Ari has been talking about how Apple’s ID moves would hit Facebook’s attribution model well before many in the industry were wise to what was coming. He and I chatted a few weeks ago for a recent episode of Next in Marketing, and he was taken aback by the fact that Facebook didn’t have something else big cooking to fix what Apple has wrought on its ad machine.
“The big x factor is, ‘what other tricks does Facebook have up its sleeve?”he said. “It seems to be now that they didn’t have anything compelling cooking.”
“They’ll need to create something new.”
Which is insane if you think about it. Because who says Facebook will come up with something that works anywhere nearly as well as it’s Apple-reliant money-printing machine has over the past decade?
We’ve seen Google bounce around potential new ad targeting models over the past few years (Floc, Fledge) as part of its broader Privacy Sandbox initiative. Credit Google for taking feedback from the broader industry, but as of now we’re left with “topics,” which sounds like something Yahoo would offer circa 1999.
Is it going to work as well as the kind of precise-ish retargeting that has dominated the web for a decade? Who knows, but I seriously doubt it. Either way, Google doesn’t need Topics to survive, because search ads are its dominant revenue stream, and even with all the ID changes, search ads are crushing it.
Facebook’s issues are existential. Meanwhile, Zuck and company are putting billions into the metaverse (which to my knowledge, has no business model, and no way of accommodating those 10 million advertisers any time soon). How likely is it that his team that is working on ad tech/attribution 2.0 won’t get distracted?
And again, even if Facebook comes up with a new, broader way of tracking which ads worked and which didn’t - one that doesn’t rely on tracking individuals - will it really work as well as the old one? And who will tell us it is or isn’t working (Facebook?)?
Even as so many of us have just expected constant magic from these companies, there are no guarantees here.
“We don’t know a lot,” said Lewine. “We don’t know about privacy sandbox, we don’t know about legislation. We’ve never given contextual a real chance. There’s a lot of unknowns.”
Of course, you have companies like the Trade Desk pushing new identifiers, which Ari isn’t so bullish on. He doubts that brands will ever be able to target a fully or mostly logged in web - thus TripleLift is building tech for publishers to help they learn more about their anonymous visitors. The idea is to build smarter ad targeting options using web surfing data and other information gathered from first party cookies - which is technically first party data.
Me, I’m not so sure. I keep harping on this, but look at the money going toward retail media. Look at Amazon’s astonishing ad growth.
Don’t forget about Walmart’s out of nowhere ad revenue growth. That’s still the most under-appreciated story in media (way more important than Applebees’ nonsense). While brands and publishers look to gather more names and email addresses, while the tech giants look to triangulate lookalikes and common interest groups - the only first part data that really seems to matter to brands are credit card numbers.