The tech giants ate your homework
You’d spent years developing that perfect organic strategy. You scried SEO sites and danced through Google’s hoops on an annual basis. Your dashboards brought you delight. Then overnight – it’s hard to say precisely on which night it happened – a few people in California pulled the rug from under you and your well of website visits began to dry up. Those AI overviews (which were forced upon us all) flipped the way search results worked, whilst customers began to chat to some strange new things called LLMs in a totally different manner to before – and the traffic that once flowed through remained trapped in a slick user interface. You suddenly found yourself having to learn a new language of AI search and having to optimise in different ways.
But no bother! You could redirect cash into paid media to make amends. Alas, you might not have realised that this budget was simply going to the bottom of the funnel – capturing the 5% of buyers who were already in-market. You had a sales target to hit and a boss who asked for leads on the hour every hour so it had a logic to it. Somewhere at the back of your mind you realised that every one of your competitors was doing the same, throwing more money at the same pool of in-market buyers and getting less value for the effort. One night – again, it was hard to tell which night – the numbers started not to add up.
How did performance marketing end up like this?
Cory Doctorow, brilliant author and a digital rights campaigner from when the internet was young, has the answer. He is famous for developing a word for why major tech platforms kept, and keep on, getting worse for users and worse for businesses. He declared it: enshittification.
This is a very simple, very deliberate cycle:
First, the tech giants make their platforms great for users. The platform solves a real problem, or creates a real benefit. It costs nothing, and builds a massive, loyal audience. Think of early Google or pre-Musk Twitter.
Then comes the pivot to corporate buyers. Once users are locked in, tech giants invite advertisers to the party. Organic reach begins to dial down, and you’re required to pay to get back in front of the people who used to follow you for free. It seems like a fair exchange for a while.
Finally, they squeeze everyone. Once both users and advertisers are dependent and have no real alternatives, the platform stops being nice. Prices go up, useful organic traffic vanishes, and any remaining value is harvested for platform shareholders.
(For search fanatics, Cory highlights precisely when it happened to Google Search.)
I have some bad news. This isn’t a temporary dip – it’s the business model working precisely as intended. B2B brands have spent fifteen years renting eyeballs from digital landlords, all of whom are now raising the rent. Renting attention is not the same as building brand equity.
There are no more quick fixes, and all of what I’m about to suggest requires research, strategic choices and effort. But there’s compelling evidence that brand – and brand strategy – once again has a significant role to play in building demand in this unhinged digital era.
Building demand takes time
Most B2B transactions are more or less a done deal long before a procurement manager picks up a phone or books a demonstration. Research from 6sense reveals that the company a buyer already favours before engaging directly with the sales process goes on to win around 80% of deals. At Hallam we refer to this as the Day One list – if you aren’t already camped inside the buyer’s mind when they’re finally ready to buy, you’re basically auditioning to be the third quote required for compliance purposes.
Furthermore, as the common 95:5 rule suggests, at any given moment, for any given category, roughly 95% of your target market is completely out-of-market. They simply do not need you yet.
This seems common sense, right? We say such things on a regular basis in our talks and articles. People get it. It feels like it needs no further explanation. And yet, for many if not most B2B brands, there remains a focus on capturing the 5% of in-market buyers, not building demand from the 95%. Which, as I pointed out in the introduction, will be increasingly ineffective over the coming weeks, months and years.
Building demand requires brand work. None of it is particularly new.
What is new, however, is that suddenly every single one of the marketing fundamentals is now doing two rather different jobs.
Marketing fundamentals shape what the buyer remembers about your brand, as they always have.
But they also shape what the machine says about your brand when the buyer asks.
Category Entry Points: Buyers now tell the machine exactly what’s happened to them
A Category Entry Point (CEP) is the moment a buyer starts thinking about your category. Not you, not your brand, but the category as a whole.
CEPs are portals to the 95% of out-of-market buyers. It is the moment you try to throw a dinner party and realise you haven’t got the right cookware – suddenly you’re in the market for a massive cast iron pan. Which cookware brands come to mind for big dinner parties? Whoever has built their brand around that idea wins my hard-earned cash. Whoever built their brand around gift giving will earn someone else’s hard earned cash. There are many moments that make people think of big iron pans.
Professor Jenni Romaniuk’s work at the Ehrenberg-Bass Institute has shown that the brands that get thought of most often are the ones linked to the most Category Entry Points, these buying moments. A handful of common ones in B2B might look like the finance director of a company resigning, an audit throwing up an unexpected problem, or a regulation changing. They are moments. They are slippery. They are hard to measure. There are likely dozens, and they differ for each industry.
They are a brilliant platform for building salience through advertising. Think Diet Coke break for one of the most iconic.
Yet for twenty years, search and the culture of SEO pretty much hid Category Entry Points from marketers.
Someone whose finance director had just walked out typed “outsourced accountants” into Google, and the real contextual story and motivation remained locked in their head. Now, hidden away in a darkened room late at night, that same person is confiding the whole story to ChatGPT. Buyers are now literally describing what has happened to them, exposing fully what brings them into the category.
There’s now early evidence that content built around those situations gets noticed in AI search. Semrush ran an experiment. Instead of writing articles around search terms, it wrote them around specific moments its buyers find themselves in, such as noticing that competitors are showing up in AI answers and they aren’t. One article was picked up by AI tools every week for more than four months. Another raised how often Semrush was mentioned. But an article on almost the same subject, published the same day, barely registered – the difference was that it was written about a topic rather than a real moment in a buyer’s week. The rules of the game had changed.
All of which points the same way as twenty years of marketing science, in which the brands that actually got remembered were the ones connected to Category Entry Points.
Now, what’s good for marketing to people is good for marketing to the machine.
Brand Positioning: If you don’t decide what you stand for, the machine will find someone who has
Positioning is the place you choose to occupy in the buyer’s mind, relative to your competitors, for the customers you want the most.
A brand position that tries to be something for everyone, isn’t brand positioning, it’s the howl of a CFO looking for more revenue. It is also not just a made up sentence from a one-hour session with leadership. It is the triangulation of speaking to your customers, benchmarking (or reverse benchmarking) your competitors and their claims to find your relative differentiation, and speaking to your internal teams, especially the product teams, to find out what you do particularly well. If you haven’t triangulated those things, you probably don’t have brand positioning.
Think of the difference between Volvo and BMW. Both are things that move on four wheels, but only one of them is firmly associated with car safety – a thing it does well, that’s relatively different to the competition and is something that certain target drivers (families, for example) really care about.
When you stand for something, people see you. They engage with you. If you stand for something effectively enough, like Volvo, it becomes the lifeblood of your outputs, from advertising to content to social and so on. And when a buyer cracks open an LLM and asks about you, your whole business gets squeezed into a sentence or two, and the machine builds that sentence from whatever it can find online – and what other people say about you seems to fill the gaps.
In Ahrefs’ study of 75,000 brands, the number of times a brand was mentioned online was the strongest predictor of whether it showed up in Google’s AI answers. The most-mentioned brands appeared up to ten times more often. Advertising spend alone, however, had only a weak link to whether AI tools mentioned a brand, which shows that you can’t always buy your way into the answer.
Think different
A June 2026 study from researchers at Trine University and Texas A&M shows what happens when there’s no perceptible difference to go on. When three of the big AI tools were shown products that were identical in every respect except the name, they recommended the well-known brand every single time (the incumbent advantage).
But the moment that a lesser-known competitor had even the smallest visible advantage by way of difference (which is to say, differentiation), that dominance collapsed. The researchers concluded that the real barrier for smaller brands wasn’t really the big brands’ reputation, for once, but rather the absence of anything that set them apart.
A very recent study from researchers at Northwestern University and Boston University seems to reinforce this. When they asked six different AI models to recommend certain brands, big names such as L.L.Bean and Craftsman often didn’t appear, so popularity was a poor predictor of who actually did show. Yet, when researchers described buyer needs using the brands’ own positioning language, those same brands suddenly appeared in more than eight out of ten recommendations. The authors concluded that their findings reinforce one of Kevin Lane Keller’s core positioning principles, which is: you get the category basics right, then build a small number of distinctive, relevant points of difference. A clear position helps the machine match you to the right buyer – just provided it’s expressed in the words buyers actually use.
Decide what you stand for, and who you’re for, and do this by having engaged with your customers. Make it specific and something you can own. Say it the same consistent way everywhere you publish, and give trade press, partners and customers good reasons to say it too. It’s proper marketing and it’s good for both humans and the machines.
Distinctiveness: When everyone Is optimising for average, standing out is even more important
AI is extremely good at producing the average – because that’s basically what it’s built to do. LLMs like ChatGPT work by predicting, one mathematically clinical word at a time, what’s most likely to come next, based on patterns derived from stealing vast amounts of human output.
Your LLM doesn’t really know what’s true, or what’s clever or original. It only knows what’s expected. If you ask it for a brand positioning statement, it will give you the kind of positioning statement that’s been written thousands of times before, which it has read thousands of times before, because that’s… well, it’s the most probable answer, and therefore a spectacularly useless one.
Now picture every B2B marketing team firing up the same tools and preparing to “create”. The same LinkedIn posts, the same case study formats, the same “people-first, tech-enabled solutions” copy spews forth. The same-looking brands saying the same things, on the same websites. Don’t even get me started on those hideous AI-generated community posters plaguing the world.
Byron Sharp, guru of Ehrenberg-Bass and of much modern marketing science, has long argued that what really matters to buyers is whether they recognise you. Whether you are distinctive. Everywhere buyers do see you in full, in ads, on LinkedIn, in video, at events, in a pitch, your brand has to be recognisable instantly and consistently, so that when your name turns up in a list of five, it already means something; it is already familiar.
Which means whenever you rely on AI to generate something for your brand, you’re likely, at a macro level, losing anything that made you distinctive in the first place, and merging with whatever your competitors are likely doing too.
So the first thing you can do to be recognisable in buying situations is to codify a handful of Distinctive Brand Assets. Yes, a logo, but more than that: key colours, a mascot, a jingle or start-up noise, a tagline if it’s been around long enough. These are cues that are unmistakably you. Find them, use them everywhere in your advertising, your content, your sales materials and your event stand, magnify them, and make it easy for customers to recall you.
Digital trust is collapsing; nobody knows what’s real anymore
The whole of digital culture has unmistakably changed in recent years. Bots, fake reviews, deepfakes, AI-generated filler posts and engagement farms have made it next to impossible to tell what’s real or not. Imperva found that in 2024, for the first time in a decade, automated traffic overtook human activity online, at 51% of all web traffic.
People have noticed. The Reuters Institute’s 2026 Digital News Report found 62% of people globally are now concerned about what’s real and what’s fake online. Every Western European market saw that concern rise. Do you think Sam Altman and Elon Musk will contribute to improving this culture? Exactly.
In B2B, trust decides deals, and if buyers can’t believe what they see online, they fall back on what they know.
So strategically you might want to consider going where trust is demonstrably higher. The evidence is consistent across countries, researchers and time.
Radio – The European Broadcasting Union’s 2025 analysis found radio is the most trusted medium in two-thirds of European countries. It’s been at or near the top of that table for more than a decade.
Print – In Ofcom’s 2025 UK research, around seven in ten users of TV, print and radio rated them well for accuracy and trust, against six in ten for online sources and four in ten for social media. Print is the one getting stronger: its trust rating has risen from six in ten in 2018 to seven in ten in 2025.
Advertising in traditional media generally – When YouGov asked Britons which advertising they trust, TV and radio came top at 53% each, followed by print at 42% and billboards at 38%. Ads on websites managed 18%. Social media managed 10%.
There’s a reason these types of media earn trust. They’re hard to fake, and they cost real money.
Trusted media doesn’t just work on its own, as it makes your digital spend work way harder. When Thinkbox analysed £1.4bn of media spend across 50 brands, TV lifted the performance of the other channels in a campaign by up to 54%, and print by up to 13%, with an average multiplier across all channels of around 8%. Radio shows the same effect on search. Radiocentre’s Online Multiplier study found listeners exposed to a brand’s radio advertising were 52% more likely to include that brand’s name in their online browsing, and a separate Radiocentre test found people were over three times more likely to search for a specific brand name if they’d heard it advertised on radio.
It isn’t complicated. People who’ve seen or heard you somewhere that they trust are obviously more likely to search for you online by name, click on your name and choose your brand.
Conclusion
Nobody here is saying tear up the performance plan. When demand for your brand arrives, you still need to catch it. But as I’ve pointed out, times are changing and the current world is an enshittified mess.
The brands that endure and thrive over the coming years will be the ones that made their own weather: they connected to the moments that sent their buyers looking, they were clear about what they stood for, they were impossible to mistake for anyone else, they showed up in trustworthy places, and they were already on the shortlist – in the buyer’s head and in the machine’s – before anyone typed a word. In other words, they thought strategically, they thought broadly, and they built brand demand.
A B2B brand strategy checklist
- Category Entry Points
☐ We’ve long-listed at least 20 real moments that send buyers into our category, in the words buyers actually use, by actually speaking to them.
☐ We’ve created a shortlist of 3-5 and checked them with sales and with customers.
☐ Our content, ads and sales materials are built around those moments, not just our category name.
☐ We know which moments we already own, and which ones our competitors own.Supports: paid, AI search.
- Brand positioning and proper differentiation
☐ We can say what we stand for, for whom, in a single sentence – with a straight face.
☐ It’s specific and provable, and not utterly empty of meaning, like “people-first, tech-enabled solutions”.
☐ We’ve based this on speaking to our customers and reverse benchmarking our competitors.
☐ We say it the same way everywhere, across our messaging, our trade press, partners and our customers repeat it.
Supports: paid, AI search. - Distinctiveness
☐ We’ve codified a handful of Distinctive Brand Assets like colours, a phrase, a character, a sound.
☐ We use them consistently, everywhere, for long enough that buyers recognise them.
☐ Our name and the way we talk would still be recognisable in a block of plain text.
☐ Nothing goes out without being checked against the use of those assets.
Supports: paid.
- Brand building beyond performance
☐ We know where our buying group reads, listens, drives and meets, offline as well as online.
☐ Some of our budget goes to trusted channels such as print, radio, outdoor and events, not just capture channels.
☐ We judge those channels by what they do to branded search and to the rest of the plan, not just last-click.
☐ We’re building demand with the 95% who aren’t buying yet, not just chasing the 5% who are.
Supports: paid, organic and AI search by building demand beyond the capture channels. The effect compounds over time.
