Don’t let AI disguise a backwards slide for agencies and brands

Are buyers seeing the full picture on Google’s Performance Max and Meta’s Advantage+ AI products?

New ‘smart’ motorways in the UK have been scrapped. Meanwhile, GM-owned Cruise’s self-driving beta hit a roadblock in January and had to halt services due to errors.

What does this tell us? Progress isn’t always one way, or bump-free. New tech needs scrutiny. It merits excitement, but not utopianism.

All of which brings us to the new products being offered by Meta and Google in the shape of Advantage+ and Performance Max respectively. Some have accused them of reducing transparency for the buyer, even as they promise greater results through AI automation – and arguably less control.

This takes us all the way back to the birth of programmatic – when real-time bidding emerged as a solution, offering greater transparency to advertisers than the ‘blind’ ad network buying that preceded it. Just as those ad networks reportedly claimed considerable fees alongside their closed shop approach, so the tech giants have been accused by some of pushing advertisers towards newer products as a way of enhancing their profits.

The second charge – which Lloyds Bank's Nic Travis makes here – is that, with Performance Max, Google is effectively ‘dumbing down’ its offerings, to be more appealing to SME advertisers, following in Meta’s footsteps. Meanwhile, larger advertisers are left out of the equation.

Those who would benefit from the customisation and functionality they were used to in the past are simply being ignored. The move arguably also allows Google to get closer to brands, removing agencies from the equation – after all, if they can’t optimise effectively, it becomes much trickier to justify their fee.

AI layoffs

Another aspect of this drive for automation – not to mention profits – is internal. Tech giants are reducing headcount in sales, and customer support – just as AI takes the reins for campaign optimisation. In short, software is providing automated support which used to come from people.

The AI layoff trend applies across Meta, Google and TikTok. As well as contradicting the old trope that tech adoption creates rather than kills jobs, there are already examples where customer service automation has adversely affected advertisers. 

As early as April last year, there were reports of major unintended overspend because of AI gone awry: “To the tune of tens – if not hundreds – of millions of dollars. Meta is still considering whether and how to issue refunds. Eight agency and brand advertisers and two Meta employees told AdExchanger they expect only partial refunds, which will come after weeks – perhaps more than a month – of campaign assessment.”

Later, according to the same title: “Despite restless clients, Meta’s ads business seems unaffected for now. In October, the social media giant predicted 25% ad revenue growth for the year, and its stock is up 164% in the past year. Plus, Meta’s data advantage and vast user base make it practically impossible for advertisers to walk away.”

AI in disguise

Next to consider are the tactics being used by tech gatekeepers to sell in their new offerings.

When Meta announced a series of apparent privacy enhancements, some commentators saw this as a fig leaf for its shift to lower transparency, and reduced control on the part of buyers. Meanwhile, in an interview with Stratechery, Om Malik described Google’s challenge as, "The curse of the cash cow. The company has shed its moonshots, built its hero consumer products and now has nothing left to do but make the advertising revenue number go up.”

And how much have the new ‘self-driving’ AI products added? Richard Kramer of Arete Research estimated during his Outlook for 2024 presentation that Advantage+ and Performance Max were responsible for $3-5bn in incremental ad spend in 2023, as they “favored their own inventory.”

As with the customer support examples, when it comes to transparency, there are cautionary tales here. Last year, independent research house Adalytics ran a series of reports exposing everything from major brand safety concerns, misreporting of placements to the targeting of ads to minors. All of which coming just as new products promise fewer insights or controls over where buyers’ ads will run.

The promise of AI around efficiency, accuracy and creativity is all massively exciting. But, at least at the current stage, human checks and balances are still required. Long-term, AI will lead to enhanced accuracy and performance across the board.

But for now, with systems still prone to hallucinations and inaccuracies, it’s still too soon for self-driving mode, not least where big budgets are involved.

By Artiom Enkov, VP insights & analytics, Nano Interactive