AI is increasingly shaping B2B buying decisions long before customers begin comparing suppliers. Sara Donnelly, founder of Beyond Echo PR, explains why brands need to understand which sources are informing those early conversations – and why the most influential media may not be the outlets with the biggest audiences.
Most buying journeys used to start with a search for a product or a supplier. Increasingly, they start somewhere else.
Someone has a problem they cannot fully identify. They sense something is wrong, but they do not yet know the answer, the name of the solution, or which companies to consider. So they ask AI.
Take Sam, for example.
Sam runs operations for a 150-room hotel. Guest satisfaction has been slipping, although not enough for anyone to point to one obvious cause. Instead, there are a dozen small problems – rooms taking too long to turn around, housekeeping teams stretched at certain times of day, and tasks being picked up in the wrong order. Enough to make people wonder what is going on.
So Sam does what a growing number of buyers now do. She asks ChatGPT.
She explains what is happening and asks why guest satisfaction might be falling when staffing levels have not changed. The answer walks her through task sequencing, room allocation, and how good hotels organise housekeeping. It suggests the problem may be less about headcount than about how work is planned and allocated.
The answer gives Sam something to think about and, more importantly, gives her a way of describing the problem.
It does not mention your company.
The shortlist isn’t where influence starts
Sam has not started shopping for ‘housekeeping software’ yet. She is still trying to understand what is wrong. But she has already been given an explanation, and that explanation will influence what she looks for next.
This situation is where the old rules of PR start to look rather different.
Tom Rudnai, who built Demand-Genius, a platform connecting website engagement with CRM data, says this is where businesses often misunderstand the role of AI in a buying journey.
“An AI like Gemini is not a directory,” he explains. “It very rarely searches for answers – about 16% of the time in a B2B journey.”
That is an important distinction. Businesses naturally think about traffic because it is something they can see and measure. AI can send people to a website, but Tom’s research suggests its more interesting role is earlier in the process.
Influence.
His research found that brand citations average around 16% across a buying journey. At the earliest stage, when someone is still trying to understand the problem, the figure falls to almost zero.
Think about what that means for Sam.
She can spend weeks learning about the problem without ever seeing your brand name. By the time she starts looking at suppliers, she may already have decided what kind of solution she thinks she needs.
The decision did not suddenly appear when she typed the category into a search box. It had been taking shape before that.
Who is teaching the machine?
For fifty years, prestige in the media was assigned in a fairly familiar way. Big circulation mattered. So did a famous masthead, a well-known journalist and the sort of logo that looked impressive sitting underneath an “As featured in” strip on a website.
There is nothing particularly wrong with that system; it just no longer provides the complete picture.
A Muck Rack study of more than a million AI citations, published in May 2026, found that Claude’s most-cited outlets have roughly half the readership of those most cited by ChatGPT. Claude also disproportionately cites trade publications and specialist sources, which traditional PR planning often treats as secondary.
The more surprising finding was the overlap between the journalists that PR teams pitch to and the journalists AI models cite. It was just 2%.
Two per cent.
If you spend most of your time deciding which journalists are influential, it is worth asking whether you are measuring influence in the right place.
The journalist with the largest audience is not necessarily the person whose work will help explain your customer’s problem. Sometimes it is the specialist who has spent ten years writing about one very particular corner of an industry. Or, a trade publication that has never made it onto the shortlist for your biggest PR campaign.
Where does the trust come from?
Research from Stacker and Scrunch, covering 944 prompt-platform combinations, found that earned media accounted for 84% of AI citations. Brand-only content appeared in 8% of answers, while trusted third-party outlets cited the same story in 34%.
A company can publish a very good article on its website and still have very little influence over the wider conversation. A respected third party explaining the same issue is doing something different. The reader brings a degree of trust to the source before they have even considered what the company has to say.
We have always worked this way, of course. A recommendation from a colleague feels different from an advert. A specialist journalist explaining an industry problem feels different from a vendor explaining why their product solves that problem.
AI makes the sources behind the explanation more important.
So what is your PR team supposed to do?
The obvious response is to produce more content. There is a certain logic to it. If people are asking more questions, give them more answers. If AI is reading content, create more content for it to read. The technology makes it possible to publish at a speed that would have seemed ridiculous a few years ago.
But there is a problem with that approach.
If everyone can produce an explanation of the same topic in seconds, the number of explanations increases, but the amount of genuinely useful information may not.
Tom has watched brands scale AI-written content aggressively. He has seen the initial lift in visibility followed by a decline, and he puts the problem rather well: “The optimisation was fine, but the differentiation had fallen to absolutely zero.”
What should you be looking for instead?
Start with the question your customer asks before they know what your product is called.
For Sam, it might be: “Why is guest satisfaction falling when my staffing levels haven’t changed?”
That question is much more revealing than “Who are the leading housekeeping software providers?”
Ask the first question and look carefully at the answer. Which publications are being drawn on? Whose research appears? Which journalists or specialists keep coming up? What information is being treated as authoritative?
That is your trust list.
And it may look rather different from the media list sitting in your PR team’s planning document.
About Sara Donnelly and Beyond Echo PR
Sara Donnelly is the founder of Beyond Echo PR, a Wigan-based consultancy specialising in B2B and technology PR for ambitious businesses across the Northern tech ecosystem.
She launched Beyond Echo after more than 20 years in the communications industry, including senior roles at Social Tech Communications, JAM and Firework PR. The agency combines strategic consultancy with what Donnelly describes as “no-nonsense consultancy and good old Northern grit”, helping technology companies build authority, sharpen their positioning and raise their visibility.
Beyond Echo PR previously sponsored the Scale-ups edition of Tech Companies to Watch at Digital City Festival 2025 and the Female Tech Entrepreneur of the Year category at the 2025 Prolific North Tech Awards.
