Where does AI search fit inside targeted account marketing? Why most B2B teams are still running them as separate jobs.

Need a builder? You don't ring the first number in the results page. You ask around. The neighbour who just had the extension done. The group chat. Whoever your sister used last year and hasn't stopped going on about since. By the time you actually pick up the phone, you've already got an opinion of who's good, and none of it came from the builder.

Buyers evaluating your business are doing something close to the same thing, just with a different neighbour. A growing share of that "asking around" now happens inside ChatGPT, Gemini, Copilot, whichever tool a given buyer prefers, often well before your team even knows a deal is in motion. The campaigns your team builds around a shortlist of target accounts still do their job. They just don't reach the room where a lot of that early opinion-forming now happens.

Which is really the point of this piece. Steering what AI tools say about you, and running tightly targeted campaigns at named accounts, have grown up as two separate jobs in most marketing teams. Buyer behaviour has already merged them into one conversation. It's worth looking at what that means for how a campaign actually gets built, not just where the budget line sits.

AI search visibility and account-based marketing are already answering the same question

Quick definitions, to help set the scene for the uninitiated. Account-based marketing, ABM, is a targeted approach to go-to-market: instead of casting wide, you pick specific accounts, run as 1:1, 1:few or 1:many, and build outreach, content and sales alignment around the buying committee inside each one. Generative engine optimisation, GEO, is newer. It's the practice of structuring, sourcing and distributing content so AI tools surface it, cite it, and represent your brand accurately when someone asks them a question in your category.

The two have mostly been run separately because they grew up in separate teams. But the buying committee ABM is built to reach has already grown past what any contact list captures. Forrester's State of Business Buying, 2026 puts the typical buying decision at 13 internal stakeholders and nine external influencers now, a number that climbs for anything complex or strategic.¹ Generative AI is one of those external influencers whether or not it's on your account plan, and it's arriving right as the channel that used to carry the conversation shrinks: Gartner expects traditional search engine volume to drop 25% by 2026 as chatbots and virtual agents take share.² The neighbours have multiplied. One of them is an AI.

Broad, 1:many campaigns can prime the answer before the campaign even launches

For 1:many programmes and broad category campaigns, the opportunity is timing. Publish the supporting articles, POV pieces and category content before launch rather than alongside it, and AI systems have time to index and start citing that material. So when buyers ask ChatGPT or Gemini about the topic after the campaign goes live, the answer already echoes what your campaign is saying, not a competitor's take, or nothing at all.

This isn't a guess. Semrush looked at content anchored around genuine, recurring category entry points, the specific situations buyers actually find themselves in, and found citation volume compounds over months on articles that address one of those situations clearly. Content without a defined entry point gets cited briefly, then drops off.³ Word of mouth compounds the same way. Say the right thing consistently and it keeps getting repeated long after you've stopped saying it. Treat pre-launch AI content the way you'd treat a press embargo: seeded early enough that the story's already circulating by the time the campaign lands.

Narrower 1:few and 1:1 tracks need the AI to already know your name

Tighter tracks are a different job. When you know exactly which handful of accounts, or which single account, you're targeting, you're not shaping a general conversation. You're making sure that when a specific buyer at a specific company checks, the answer holds up. That means optimising for the exact entry points that buying committee is likely to search, not the broad category terms a 1:many campaign would chase.

It matters more than it might look like it should, because buyers don't trust AI answers blindly. They use them as a starting point, then go looking for confirmation. Forrester found buyers actively seek validation from trusted sources precisely because AI search often delivers incomplete or unreliable answers on its own.¹ A 1:1 track that optimises every human-facing channel, email, LinkedIn, direct mail, but ignores what a named buyer finds when they check ChatGPT themselves, has left one of the neighbours they're going to ask unbriefed.

This is essentially PR wearing a new coat

Step back and the mechanics look familiar. What earns a citation from an AI system, credible third-party coverage, clear expertise, being the source other people point to, is the same thing that's always earned a mention from a journalist. GEO isn't a new discipline bolted onto marketing ops. It's PR's job description, with a new reader.

Leave GEO with the technical SEO team alone and it gets treated as a formatting exercise: headers, schema markup, structured answers. Give it to the people already skilled at getting third parties to say the right thing about you unprompted, and it gets treated as what it actually is, reputation management for an audience that happens to be a language model. It's the same job as briefing the neighbour who talks to everyone else on the street, just for an audience of one.

The publishers your buyers' AI trusts matter more than your own site

This is where ABM, PR and GEO genuinely meet, in the publishers. Even when an AI system is answering a question specifically about your brand, a substantial share of what it cites doesn't come from your own site. Profound looked at citation sources and found that on brand-specific prompts, 27 to 46% of citations still come from outside the brand, depending on the platform. Your own domain is never the whole story, even on your own name.⁴

Some outlets have simply got better at being the source AI systems reach for. Semrush tracked citation activity across ChatGPT, Google AI Mode and Perplexity over three months and found Forbes among the biggest winners, doubling its citation share on ChatGPT alone in the period studied.⁵ Not every publisher carries that weight, and the ones that do vary by category. Working out which publishers are actually driving citations for your specific entry points, rather than assuming one big-name helps everywhere, is what turns publisher partnerships into a targeted extension of the ABM programme itself.

This is the analysis we run for clients before recommending where to invest. Invera’s analysis mapped the citation landscape for a digital experience platform brand and it turned up 1,602 separate domains cited, with the top 50 accounting for over half of total citations. Within that top 50, earned and editorial sources (analyst platforms, review sites, trade press) made up to a quarter of the citation share, separate from the vendor and competitor domains.

A category entry point approach to those partnerships does three things:

• It matches the source to the situation. The specialist outlet a buyer's AI reaches for when researching supply chain risk isn't the one it reaches for on cybersecurity, so a single "get featured everywhere" plan misses most of the value.
• It compounds rather than expires. Citations anchored to a genuine, recurring entry point keep being pulled into answers for months, unlike a one-off press hit.
• It extends the campaign, not just the brand. A publisher relationship built around your ABM programme's specific entry points reinforces the exact conversations your target accounts are already having with AI.

Plan ABM, PR and third-party publishers together like this and you're not just extending the campaign's reach into another channel. You're reinforcing what the AI has already learned, with higher citations landing across a content ecosystem you don't fully own but can absolutely influence.

This isn't a one-off fix

Getting featured in Forbes once isn't the same as being the name people expect to see there. Word of mouth, human or AI-mediated, decays if you stop tending it. Citation patterns shift, entry points change, and the publishers AI systems trust today won't necessarily be the ones they trust in a year. Treat this as ongoing account intelligence, reviewed alongside the rest of the ABM programme, not a project marked complete and filed away.

Go back to the builder. You could pitch brilliantly, turn up on time, quote fairly, and still lose the job because the person two doors down had already told them not to bother. Your account campaigns might be reaching everyone on the buying committee's contact list. Worth checking whether they're reaching what those people hear when you're not the one talking.

Sources: ¹Forrester, 2026, ²Gartner, 2024, ³Semrush, 2026, ⁴Profound, 2026, ⁵Semrush, 2025