I think a lot of people are still looking at AI search the wrong way.
They obsess over discovery metrics — mentions, citations, impressions, visibility screenshots, all the vanity stuff that looks good in a sales deck. That’s fine, but it misses the part that actually matters: does the machine help you win the deal?
That’s why this Uniplay example is interesting.
They secured a £23,400 school project even though they were the most expensive quote out of five. On paper, that should have made them less likely to win. In reality, the buyer used ChatGPT to help evaluate the options, and the AI recommended Uniplay as the best value, not the cheapest.
That distinction matters.
A lot of SEO people still act like the game ends once you get surfaced. It doesn’t. In plenty of buying journeys now, especially where multiple quotes, PDFs, specs, reviews, and case studies are involved, the real game is happening at the decision stage.
And this is where most businesses are underprepared.
In this case, the AI wasn’t just looking at price. It was looking at the full picture:
- whether the quote actually matched the school’s requirements
- whether the specification was stronger
- whether testimonials reduced risk
- whether third-party reviews and trust signals backed up the claims
- whether the company looked like a safer bet than cheaper alternatives
That’s exactly how a rational buyer should think, and increasingly it’s how AI systems summarise the options for them.
The important bit here is that Uniplay didn’t win because they were cheaper. They won because the AI could see they were offering something better aligned to the brief.
I’ve seen this for years outside AI as well. Businesses lose jobs because buyers compare two offers as if they’re equivalent when they clearly aren’t. It’s the classic Ferrari vs Fiesta problem. One quote includes the right deliverables, the right quality level, and the right fit. The other is just cheaper. If the buyer can’t properly evaluate the difference, price usually wins.
What’s changing now is that AI can sometimes do a better job than the buyer at spotting those differences.
That should be a wake-up call.
If your site, brand, and off-site reputation don’t clearly communicate why your offer is the safer or better-fit choice, you’re leaving the decision layer open for someone else to own.
And no, this isn’t just “SEO with a new label.” Some of the GEO/AEO crowd are already turning this into buzzword soup. Same old snake oil, new acronym. But the underlying shift is real.
What I take from this is pretty simple:
AI optimisation isn’t just top-of-funnel anymore. It’s bottom-of-funnel influence.
If a buyer uploads quotes into ChatGPT and asks, “Which supplier should I choose?”, the model is going to look for evidence. Not vibes. Not your homepage tagline. Evidence.
That means your business needs:
- clear specifications
- strong testimonials
- credible case studies
- third-party reviews
- external trust signals
- consistent proof that you’ve delivered similar work before
Even older testimonials can help because they show trading history and reduce perceived risk. That’s the sort of thing many SEOs overlook because it doesn’t fit neatly into a ranking report.
But buyers care about risk. AI summarises risk. And the company that looks safest often wins.
That’s why I think “decision engine optimisation” is probably the more useful framing here. Discovery is only half the story. Being recommended at the point of decision is where the money is.
And honestly, that’s the kind of case study I want to see more of.
Not another post about impressions going up 38%. Not another chest-beating thread about “brand mentions in AI overviews.” Show me the revenue impact. Show me the deals that got closed. Show me where AI changed the buyer’s final decision.
Because that’s when this stops being a novelty and starts being commercially serious.
My takeaway: if your SEO strategy still ends at the click, you’re behind. The next layer is making sure your business is the one AI systems can justify recommending when the buyer asks who offers the best value.
That’s a very different game — and in my opinion, a far more important one.