Something Inc.Schedule a free consultation
ANALYTICS

Paid search vs SEO when AI Overviews contradict your ads

When Google's AI Overview names a different brand than the one paying for the ad, paid search vs SEO stops being a philosophical debate and starts being a line item. Here's how to close the gap with data you already own.

TTTyler TruffiManaging Partner · AUG 9, 2026 · 10 MIN READ
TL;DR · 60 SECONDSA brand can now pay for a search ad and watch Google's AI Overview recommend a competitor in the same result. Search Engine Land documented it in early August: an Eco Plumbers ad running next to an AI Overview that named Roto-Rooter and Amanda Plumbing instead, and a phrasing change that flipped the AI Overview's brand pick entirely on a second query. The fix isn't quitting paid search or panic-building a separate GEO program from scratch. It's coordinating the two, starting with the search-term data your paid team already owns.

Paid search and SEO used to run in separate lanes: one bought placement, the other earned it, and neither showed up inside the other's result. AI Overviews just erased that boundary. The same query can now serve your ad and an AI-generated answer that names your competitor, in the same result, in the same ten seconds.

That's the real stakes behind the paid search vs SEO question in August 2026. Two pieces published the same day by Search Engine Land describe opposite halves of the same problem. Sarah Stemen documents AI Overviews actively contradicting the paid ad sitting next to them. Heather Brousell argues the fix is hiding inside data most paid search teams already collect and mostly ignore. Read together, they say something neither says alone: paid and GEO aren't separate budgets with separate owners anymore. Running them apart is what's costing you the click.

For most of the last decade, the assumption held up fine. A Google Ads campaign and an organic ranking were competing for the same eyeballs but not for the same decision. You paid to guarantee a placement above the fold, and the algorithm handled everything below it. If the ad converted, it didn't matter much what an unpaid result three positions down was arguing. The two channels reported to different budgets, different KPIs, and often different people on different floors, and that division of labor worked fine because the two surfaces stayed genuinely separate.

Search Engine Land's report on AI Overviews contradicting ads is the clearest evidence yet that the separation is gone. An AI Overview isn't a third blue link competing for attention below your ad. It's a synthesized answer that names a specific brand, and Google's own model can pick a different brand than the one that just paid to be there. When that happens, the ad and the AI Overview aren't two channels sharing a result page anymore. They're competing for the same click, and only one of them costs money on a per-impression basis.

AI Overviews aren't a niche surface anymore either. Our own citation tracking has found them appearing on 43% of the searches we monitor, across categories that carry real paid budgets, not just informational long-tail queries where nobody was bidding much to begin with. Any team still treating paid search vs SEO as a budget-allocation question, rather than a coordination problem, is optimizing for a search results page that stopped existing.

Two real examples of AI Overviews contradicting the ad

Stemen's reporting includes a documented case that makes the risk concrete instead of theoretical. A user searched "plumber." Eco Plumbers had the paid ad. The AI Overview above it recommended Roto-Rooter and Amanda Plumbing instead, and left Eco Plumbers out of the generated answer entirely. Eco Plumbers was still paying for the impression. It just wasn't the brand the AI Overview decided to vouch for.

0
mentions of the paying advertiser, Eco Plumbers, inside the AI Overview shown on its own "plumber" ad
2
different AI Overview brand picks for "sweatshirts for anxiety" vs. "what is an anxiety sweatshirt" — same intent, different winner

We've written before about the same failure mode from the organic side, where an AI Overview cites a brand and recommends a competitor in the same breath. The plumber case is that pattern with a dollar amount attached, because the brand being passed over wasn't just optimizing content. It was buying the placement outright.

The second example is more unsettling because it has nothing to do with bid strategy or match type. Stemen found that "sweatshirts for anxiety" and "what is an anxiety sweatshirt" — two phrasings of the same underlying intent — produced different AI Overview brand picks. Same shopper, same need, same moment in the funnel, different winner depending on how the words landed. That's not a targeting problem a better keyword match fixes. It's an AI Overview accuracy problem sitting one layer above your ad account entirely, and no bid adjustment reaches it. Whatever a Google AI Overview decides to synthesize from a given phrasing, right now, is largely outside an advertiser's control.

WHY THIS IS A BUDGET PROBLEM, NOT A UX PROBLEMAd impressions stay stable while the AI Overview intercepts the click. You keep paying the auction price for the eyeballs while a different, unpaid brand collects the decision. Stemen's piece warns that shows up downstream as falling CTR, Quality Score erosion, and rising eCPCs — months before anyone traces the symptom back to its actual cause.

The paid search data you're already sitting on for GEO

Brousell's companion piece, published the same day, argues the fix doesn't require a brand-new content program built from nothing. It's sitting inside the paid search account most teams already run. Search-term reports show the exact conversational language real customers use, not the keyword you bid on but the actual phrase they typed into the box. High-performing ad copy has already survived thousands of impressions of split-testing against real buyer intent. Product feed and Shopping data carry structured specs that a GEO content team usually has to build by hand from scratch. None of that is GEO content yet. All of it is GEO input, already paid for, already validated against real conversion data.

Already validated
Search-term reportsThe literal phrases customers type, not the keywords you bid on. Pull the ones actually converting and turn them into headings and direct-answer paragraphs — the exact phrasing an AI Overview or ChatGPT is already trying to match against your page.
Already tested
High-performing ad copyEvery headline and description that survived split-testing already states your differentiation in plain, buyer-tested language. Rework the winners into comparison claims a model can lift cleanly and attribute correctly.
Already structured
Shopping and product-feed dataStructured specs, pricing, and attributes you maintain for Shopping ads are the same structured facts an engine needs to cite you accurately instead of guessing at a competitor's numbers.

Comparison content already earns the largest single share of AI citations we track, more than any other format, because it answers the buyer's first real question before they've picked a brand to search for by name. Ad copy that's already proven itself against real buyers is some of the fastest raw material for building it — you're not guessing at what resonates, you're recycling what a split test already confirmed.

Run the two pieces together and three postures fall out of them, not two. A team can run paid search alone and treat AI Overviews as somebody else's problem. It can run SEO/GEO alone and treat paid as a separate budget with a separate owner and a separate dashboard. Or it can coordinate the two — feeding paid search-term data into GEO content, and using citation monitoring to flag exactly where paid spend is being quietly undercut by an unpaid answer. Only one of these three holds up now that AI Overviews can name a different brand than the one paying for the click.

POSTUREAI OVERVIEW CONTRADICTION RISKWASTED AD SPEND EXPOSURESPEED TO ACT ON SEARCH-TERM DATA
Paid search aloneHigh — no visibility into what the AI Overview shows for your own paid termsHigh — keeps paying for impressions an unpaid competitor is winningN/A — the data sits in Ads, unused beyond bid management
SEO/GEO aloneMedium — content can earn citations, but nobody is watching the paid terms the team doesn't ownLow direct spend risk, but blind to where paid dollars are actively leakingSlow — the GEO team has no access to search-term reports, starts content from zero
Coordinated paid + GEOLow — one team checks AI Overview output against both the ad and the citation positionLow — spend gets reallocated, or the ad gets pulled, the moment a term is being lost to an uncited AI OverviewFast — search-term data flows directly into GEO content as validated raw material

Coordinated wins every column, and it isn't close. The interesting part is that it's also the cheapest posture to adopt, because it doesn't require new tooling or new headcount to start. It requires the person who owns your Google Ads program and the person who owns your citation tracking to look at the same twenty search terms in the same meeting.

How to coordinate paid and GEO starting this week

This doesn't need a quarter-long initiative or a reorg. It needs a recurring habit that ties two reports together that currently live in two different tools, owned by two different people who rarely compare notes.

1Do this before your next planning meetingPull your top 20 converting paid search terms
THE MOVES
Export the highest-converting search terms from your Google Ads program for the last 90 days — not the keywords you bid on, the terms that actually triggered your ad
Sort by conversion volume, not spend, so the list reflects what buyers actually type and act on
DONE WHENYou have 20 real customer phrases, ranked by proven conversion, sitting in one sheet.
2Same weekCheck what a Google AI Overview shows for each one
THE MOVES
Run each of the 20 terms manually and record whether an AI Overview appears, which brand it names, and whether that brand is you
Flag any term where a competitor is named while you're the one paying for the ad next to it
DONE WHENYou know exactly which paid terms are being undercut by an uncited AI Overview, by name, not by guess.
3Following two weeksTurn the flagged terms into GEO content
THE MOVES
Rebuild the flagged terms' language into direct-answer content and comparison pages built to be cited, not just to rank
Route the winning ad copy for those terms into the new page's claims, since it's already buyer-tested language rather than a guess at what resonates
DONE WHENThe pages most likely to close the AI Overview gap are live, built from language you already know converts.

Fold the results into whatever dashboard already ties your channels to revenue. If attributing pipeline to organic and AI-cited traffic is still split across two disconnected reports, this is the same blind spot with a new symptom: two systems, two owners, nobody reconciling them on a weekly cadence. A generative engine optimization program that starts from validated paid-search language ships citation-worthy pages in weeks, not a quarter, because the hardest part of GEO content, knowing what buyers actually ask, is already sitting in an export you haven't looked at yet.

DO THIS NEXTExport your top 20 converting paid search terms from Google Ads today. Run each one manually and write down whether an AI Overview appears, which brand it names, and whether that's you or a competitor. That one spreadsheet is the entire business case for coordinating paid and GEO — and it's the same spreadsheet Brousell's framework and Stemen's warning are both pointing at, from opposite directions, on the same day.

See where you are cited today

A free snapshot audit of your rankings and AI citations before we ever talk.

TT
Tyler TruffiMANAGING PARTNER, SOMETHING INC.

Tyler leads work at the intersection of SEO and generative engines at Something Inc., helping B2B brands get ranked and cited across every major AI engine.

Free consultation

Let us be the last SEO agency you ever work with

A 30 minute call and a free audit of your SEO and GEO position. You keep the findings either way.