The Line Item That Slips Past Google’s Auction

  • AI
  • November 16, 2026
  • 0 Comments

Fahad Difcha makes his living moving other companies’ advertising money, and for most of his career that meant one destination. As chief executive of AccuraCast, an agency that buys media for a roster of brands, he treated the search auction as a toll paid at the top of every plan. This year the toll has competition, and the competition arrives in a chat box.

His verdict is short. “Real competitors to Google have appeared in this market this year, such as ChatGPT’s ad business and Meta,” he said. “We will reallocate budgets according to performance.” A man paid to place bets is describing a bet he can now place somewhere else, which is the sentence Google’s advertising operation has spent a year trying to make impossible.

The Levers an Advertiser Never Sees

Google’s answer runs through its own machinery. The company has pushed AI into the parts of search advertising a client never touches: deciding which queries are commercial enough to carry an ad, choosing the landing page a click will open, and generating several versions of the same copy so the system can test them against each other.

Each of those used to be a lever a buyer pulled by hand. Keyword lists, negative matches, copy rotations, the landing page built for a campaign; the craft of media buying was partly the craft of arguing with a machine about those choices. Handing them to a model is sold as efficiency, and for a small advertiser with no analyst it probably is. For an agency that charges for judgment, it moves the job.

Behind the public pitch the company has been working the other end of the problem. It is pursuing partnerships and weighing potential acquisitions aimed at making advertising inside AI features more useful, which is an admission that the ad has to survive a format where the answer arrives before the link does.

That is the structural worry, and it does not depend on any single quarter. Search advertising was built on a shelf: a page of blue links, a few paid slots above them, a surface that exists to be scanned. A generated answer removes the shelf. The better the answer, the fewer reasons to click, and every click that does not happen is inventory nobody sold.

Six Advertisers, One Direction

Evidence that the shelf is wobbling comes from the buy side. Six advertisers and agencies say dissatisfaction has already pushed budgets toward other platforms, Meta among them. What they describe is a change in where the money lands rather than a verdict on whether Google’s systems work.

Meta’s counter-offer has a different shape. It also sells automated creative tools built to drive conversions, but its newer AI products, the Muse agent among them, do not carry advertising that competes with the main feed. Ads attach to short video and social content instead, interrupting something a person chose to watch rather than an answer a person asked for.

For a brand, that is a legible trade. Feed advertising has been measured for a decade, the formats are familiar, and the audience is not in the middle of a task. Search carried the opposite virtue: intent, declared in a box. The question this year is whether intent still belongs to the box or has moved into the conversation, where an ad has to be woven into prose instead of parked beside a link.

Difcha’s answer is procedural. He compares vendors on the same sheet, reads the conversion numbers and moves the line item. That is what an agency does when a market stops behaving like a monopoly, and it is also what makes the shift awkward for Google to reverse: once a budget moves, it has to earn its way back with results rather than with a keynote.

The Cost of a Better Answer

Google’s difficulty is that its strongest product and its most profitable one pull in opposite directions. A model that answers cleanly is a better search engine. It is a worse place to put an ad, because the surface where ads live is a list of things a person still has to go and read.

The company’s public response has been to argue that AI expands the opportunity, and there is a version of that story in which it does. Sharper targeting, cheaper creative, landing pages that convert more often: all of it raises the value of a click. Advertising has absorbed every previous change in the plumbing, from mobile screens to video, by finding a new unit to sell.

The difference is who controls the unit. On a results page, Google owns the layout. Inside a generated answer, the ad competes with the text for the same few lines, and the reader’s patience is the scarce resource. Every advertiser who has watched a format degrade knows what happens to prices when the audience learns to skip.

There is also the matter of what the six advertisers can compare. A chat product that serves ads reports clicks and conversions on its own terms, and a buyer who cannot put two numbers side by side falls back on the one he trusts. Google’s advantage for twenty years has been that its numbers were the numbers everybody else was measured against.

Back at the Sheet

Difcha will make the call the way he always has, one line at a time, and a line is not permanent. Budgets that move on performance can move back on performance, which is the part of his sentence Google can still work with.

What changed is the default. For two decades the first question in a media plan was how much of it belonged to search. This year the question is whether it belongs there at all, and Difcha answers it with a spreadsheet that has more than one column.

Related Posts

  • November 20, 2026
  • 17 views
Sandisk Completes Eight-for-One Split as NAND Outlook Darkens

Eight for one. Sandisk’s stock split took effect on Nov. 19, and the shares began trading on the adjusted basis on Nov. 20, a change that divides the price by…

  • November 20, 2026
  • 16 views
A Lawmaker’s Privacy Suit Puts Model Design on Trial

For three years British courts have sidestepped the question of whether the company that builds an AI model answers for what its users produce with it. A written defence filed…