When ChatGPT Recommends Your Competitor

Last year the question was whether AI search would matter. This year it does. ChatGPT, Perplexity, Gemini and Google's AI Overviews now write the answer for the buyer: they name specific brands, build the shortlist, say which suppliers to consider. A new shelf has appeared inside the answer, the recommendation surface. The brand named on it gets the deal. The brand absent from it is invisible.

Most marketing teams have not yet priced the competitive implication. Lindsay Boyajian Hagan, VP Marketing and Co-Head of Revenue at Conductor, puts it plainly. If a brand has no presence in the sources AI engines read, a competitor's name appears instead.

Chris Bagnall, CEO of Transmission, the world's largest independent B2B agency, names the second-order effect. No board signs off on a supplier they have never heard of, however well it performs in evaluation. By the time the buying group reaches the boardroom, the engine has already narrowed the shortlist.

For senior leaders this is a competitive intelligence problem, not a content one. AI brand citation is now a leading indicator of who makes the buyer's shortlist, and most dashboards do not show it.

The six operators interviewed here converge on one point: AI visibility has become an audit function, a budget line and a board metric. A 2026 plan without a line for the work of getting named funds only part of how the shortlist now forms.

The buyer is arriving with the engine's answer in hand

Chris Bagnall, CEO of Transmission, runs an agency that helps the world's largest B2B brands win consideration. The behaviour he is now seeing in his own buying decisions is the same one his clients are starting to see in theirs. When the agency was sourcing materials for a recent project, an AI engine generated the shortlist.

The mechanics matter. The engine does not present every supplier it could find. It presents a small number, weighted by what it can confidently cite. The brand cited has been introduced to the buyer. The brand not cited has been excluded, however well it would have performed in a fair evaluation.

Bagnall's second observation lands at boardroom level. Even when the formal evaluation happens later, the supplier name has to be recognisable enough for the buying committee to sign off on it.

The two observations combine into a single competitive picture. The engine narrows the consideration set. The board narrows it again on familiarity. A brand absent from both is no longer competing for the deal. By the time procurement opens, the question has already been decided, just somewhere the supplier's own sales team was not in the room to influence.

The new shelf inside the answer layer

The previous piece in this series set out the answer layer: the space above the blue link where the engine writes the buyer's answer without sending the click anywhere. Inside that answer is a small, finite space where specific brands get named. Call it the recommendation surface. It is the new shelf, and the answer layer is the new aisle.

Lindsay Boyajian Hagan, VP Marketing and Co-Head of Revenue at Conductor, watches this competitive picture play out across the enterprise client base her platform serves. Her articulation of the consequence is the clearest in the field.

The legacy SEO competitive picture had two clean variables: how the brand ranked, and what the buyer chose to click. Rank was the contest. Click was the outcome. The recommendation surface has a different shape. There is no list to climb. There is an answer being written, and the question for the brand is whether it will be named inside that answer. If the engine names a competitor instead, the buyer never sees the alternative. There is no second position to fight from. The contest is binary at the level of the individual prompt, and it compounds at the level of the prompt set.

The compounding is the part the boardroom has not yet priced. A competitor named in eighty per cent of the prompts a buyer in the category would naturally type is, in effect, the default answer to most of those prompts. Default answers compound into default consideration. Default consideration compounds into default shortlists. Default shortlists compound into pipeline that did not exist a year earlier, on the other side of the desk.

The audit set the engine is using

The engines do not pick names at random. They build answers from the sources they can read with confidence. Martin Kihn, SVP of Strategy at Salesforce Marketing Cloud and one of the most-published voices on marketing technology, frames the underlying discipline.

Underneath sits a practical signal set: topical authority on the brand's own site, mention frequency in third-party publications, structured machine-readable information, peer-driven content on Reddit and LinkedIn, review sentiment, and consistency of voice across owned channels.

The marketing function already produces most of these. What is new is that they have stopped being content-marketing outputs and become the inputs the engine reads to decide which brand to name.

Ashley Bassman, Program Director of Product Marketing for IBM's data and AI portfolio, sees the same signal set from inside an enterprise marketing function. Her observation is that the data already exists.

Bassman's point is that this needs no new tooling. The brand-health signals marketing has tracked for years, NPS, third-party reviews, share of voice on the platforms buyers gather on, are now the inputs the engine reads. They were communications metrics. They are now discovery metrics. The board pack that left them out last year needs them in this one.

A boring brand cannot be recommended

The harder problem sits underneath the signal set. The engine can only recommend a brand it has heard distinctive things about. Paul Cash, CEO of Rooster Punk and an adviser on brand-led growth, has spent three decades watching B2B brands optimise themselves into invisibility.

The engine is, in effect, a fast reader of everything people have already said about the brand. Where the source set is thin, generic or interchangeable, it has little to cite. Where the brand has built a distinctive point of view and been quoted and reviewed in distinctive terms, it has plenty. The brands the engine names are usually the ones that spent the previous five years giving the market something specific to say about them.

This is the part of the AEO investment case that gets missed when it stays inside the content team. The team can build the structured pages and the expert answers. They cannot, on their own, give the brand a distinctive position to express.

That sits with the CMO and, increasingly, the board. A budget that funds the AEO line without funding the brand authority underneath it pays for the audit but not the thing the audit needs to find.

The competitive intelligence reframe

The mechanics of running the audit are no longer difficult. Renaye Edwards, Global COO and MD at Ammunition, frames the influence channels marketing now has to operate across.

The practical audit is straightforward. Map the fifty to one hundred prompts a buyer in the category would type into ChatGPT, Perplexity, Gemini and Google AI Overviews. Run each one. Record which brands are named, in which order, and from which sources. Set the brand against three to five primary competitors across the prompt set. The competitive picture inside the recommendation surface becomes visible at this point, often for the first time.

The picture is rarely flattering on the first read. Boyajian Hagan, who watches this run across Conductor's enterprise clients, finds most brands are being out-cited by competitors they did not consider primary rivals, on sources they did not consider primary channels. The audit re-maps the competitive field. It belongs in front of the board.

Run once, the audit is a snapshot. Run quarterly, it is competitive intelligence. Citation share moves quarter to quarter as engines update, content lands and competitors invest.

That movement is a leading indicator of shortlist inclusion, six to twelve months before it shows up in pipeline. It is exactly the kind of early signal boards have always asked marketing for and rarely been given.

What this means before the next planning cycle

Three implications sit at board level. They do not require the marketing function to abandon what it is already doing. They require the function to add what is missing.

First, the citation audit. The function needs a documented quarterly picture of its citation share against three to five primary competitors, across the prompt sets buyers actually use. The tooling exists; AEO platforms like Conductor provide it. The cost is modest against the pipeline these answers are shaping.

Second, the source investment. The audit reveals where the engines read the brand from, and where they do not. The investment case follows: visible expert presence on Reddit, LinkedIn and the platforms relevant to the category, topical authority on the brand's own site, structured machine-readable information, and the distinctive brand position that makes the brand quotable in the first place. Each is a line on the 2026 plan, now accountable to a measurable outcome inside the recommendation surface.

Third, the board pack promotion. Citation share, sentiment, mention rate and AI-sourced referral traffic should sit alongside organic sessions in the dashboard the board sees. The data is in reach; elevating it is the marketing leader's call. Bring it before the CFO asks, rather than reporting on yesterday's funnel while a competitor wins today's shortlist.

The boardroom will recognise the wider implication at once. AI brand citation is now competitive infrastructure. Audit the footprint, fund the sources, promote the measurement, and the brand is doing the work of getting named. Leave it, and the engine and the competitor write the shortlist together. The buyer who asks an engine for a shortlist is being given one. The brand absent from it is absent from the deal.

Contributing Practitioners

The voices behind this piece

This analysis is built from long-form interviews conducted on The Business of Marketing podcast with six senior practitioners on AI brand visibility, the recommendation surface inside generative AI answers, citation auditing, brand authority and the competitive intelligence reframe every senior marketing leader now needs.

Chris Bagnall

Chief Executive Officer

Transmission

Martin Kihn

SVP Strategy

Salesforce Marketing Cloud

Paul Cash

Chief Executive Officer

Rooster Punk

Ashley Bassman

Program Director, Product Marketing

IBM WatsonX

Renaye Edwards

Global COO & MD

Ammunition

Lindsay Boyajian Hagan

VP Marketing & Co-Head Revenue

Conductor