We are missing the true value of SEO – and AI search is about to expose that.
Right now, half the industry is obsessing over how to measure citations in LLMs. Everyone is building trackers and dashboards: which model mentioned us, in which mode, how often, can we link that back to a URL, can we somehow squeeze this into attribution? At the same time, there’s a meta-discussion about whether any of it is reliable anyway because answers are personalised and volatile.
All very clever. All very familiar. And it quietly repeats the same mistake SEOs have made for years.
We are still treating SEO as “the thing that brings traffic and sales” instead of seeing it for what it really is: an ongoing market and customer intelligence engine.
How we sold SEO short
For a very long time, most SEOs have told clients and C-level something along the lines of:
“Good SEO brings you more visibility, more traffic and more conversions.”
We turned that into a neat stack of numbers: impressions, clicks, rankings, visibility scores, maybe some share-of-voice charts, and now AI Overview presence and LLM citations. That’s the story: the arrow went up, therefore SEO is working, therefore we’re adding value.
The problem is that this quietly narrows SEO down to a channel. A slightly mysterious, algorithm-driven channel, but still just a channel. Input, output, ROI. The same way you’d talk about paid search or social.
And once you frame SEO that way, you also invite the obvious follow-up question: “Okay, but how much revenue did organic bring in this quarter?” That’s where we start gluing together attribution models, last-click vs. assisted conversions, and now AI-visible vs. not-AI-visible, and pretend the resulting spreadsheet is reality.
It’s not. It’s a useful approximation at best. At worst, it hides the real value.
Traffic and sales: nice, but not the whole story
Don’t get me wrong: traffic, leads and sales matter. If your work never moves those numbers, you’ve got a different problem.
But we all know the sad little secret of “SEO success”:
You can drive tons of top-of-funnel visits that are so far away from the actual core offering of the business they might as well be on another planet. Lovely graphs. Zero impact. A blog post that ranks for some broad informational query might be fantastic for visibility, but if those people were never remotely in-market for what the company actually sells, what did we really contribute?
In many cases, the honest answer is: almost nothing.
And even when we do a better job – when we work closer to the money, track product revenue, MQLs and SQLs from organic – we’re still mostly stuck in attribution land. That’s already a step up from pure “visibility”, but it still treats SEO as yet another marketing pipeline that’s supposed to push people through with minimum leakage.
It’s closer to the truth, but it’s not the truth.
SEO is not just a sales channel
Here’s the uncomfortable bit:
SEO is not a sales channel.
Yes, sales and leads are outcomes. They are the cherry on top of the cake. But the cake itself is something else entirely.
SEO – and now AI / LLM search – is the place where you see what people really think, want, worry about and compare. It is the ongoing conversation between your market and your brand. All those queries, all those SERP patterns, all those LLM answers: that’s raw, messy, honest feedback.
Every day, the data we look at can tell us things like:
- what people believe your product does (and doesn’t do),
- what they are unsure about, confused by or actively sceptical of,
- which competitors they hold you up against and why those sometimes look more appealing.
But we rarely present this as “the product” of SEO. We treat it as background noise and only cherry-pick the pieces that help explain the traffic and conversion curves.
That’s where we’re underselling ourselves.
The myth of the perfect persona
Most companies have personas: nice clean documents with names, job titles, pains and goals. In reality, those personas are often a mix of wishful thinking and workshop fiction.
They are fine as a starting point. The problem is when they become static. If you don’t constantly adjust them with what you see in search data, content consumption and now LLM outputs, they slowly drift away from reality. You start optimising content and funnels for people who don’t actually exist, while ignoring the messy mix of segments who do.
Search and LLM data show you the real audience:
- the words they use rather than the words in the brand deck,
- the use cases they actually care about,
- the objections they keep repeating,
- and the weird edge cases you’d never have invented in a persona workshop.
That’s not just “SEO input”. That’s information every product manager, marketing leader and salesperson should be begging you for.
LLMs make the intelligence layer bigger – and more obvious
Now we add LLMs and AI search into the mix. Everyone panics about losing “classic” organic visibility while at the same time trying to bolt new metrics on top: AI traffic, citation share, answer mentions, whatever.
But what LLMs really do is turn the volume up on the intelligence layer that was already there.
When a model answers questions like “What’s the best tool for X if I’m a beginner?” or “Which ERP is easiest to implement for SMEs?” it has to pick brands, features and trade-offs. That answer might not be perfectly transparent, and it might be different from user to user – but if you look at it over time, it still tells you a lot:
- how the outside world sees your positioning,
- which strengths are associated with you,
- which weaknesses keep coming up,
- and in which scenarios you’re not even in the conversation.
No, this won’t give you a clean attribution path. But it will give you patterns. And patterns are exactly what leadership needs to make decisions.
If we insist on reducing all of that to a single metric – “LLM mentions this month” – we’re missing the point again.
Who should actually be using SEO and LLM data?
One of the biggest mistakes we make as SEOs is treating our insights as if they only concern the marketing department. As long as our reports go to the “performance” team and maybe the CMO, we’ve already limited the impact.
The truth is: SEO and LLM visibility data belongs on the desks of:
- Product and R&D, to see what features people expect, where they feel let down, which use cases look promising.
- Customer service and success, to understand what’s unclear, what keeps breaking, which promises in marketing create headaches later.
- Sales, to know the real objections, the real competitor comparisons, and the phrases customers actually use when they talk about their problems.
- C-level, to see how the category is evolving and where risk and opportunity are emerging.
If your monthly SEO report doesn’t spark any discussion outside the marketing bubble, that’s not proof that SEO is “just a channel”. It’s a sign that we’re reporting the wrong things.
Change the story you tell about SEO
Most SEO reports follow the same three-act structure:
- Here’s what we did.
- Here’s how traffic and rankings changed.
- Here’s how many conversions or leads we can attribute.
Useful, but incomplete.
We need to add another layer: what we actually learned from all this data and what the business should do with that knowledge.
That might look like:
- a short narrative about new questions users are asking,
- observations on how people talk about competitors,
- recurring concerns or misconceptions,
- and a clear suggestion for what product, sales or service should change in response.
This is where SEO stops being a line item under “acquisition channels” and starts being something more dangerous: a source of uncomfortable truth about the gap between how companies see themselves and how the market sees them.
That’s also where our value spikes.
From channel operator to strategic partner
If we continue to sell ourselves as “the folks who get you free traffic from Google (and now maybe some AI tools)”, we shouldn’t be surprised when:
- the value of our work is questioned,
- budgets get squeezed,
- and our seat at the table disappears when things get tough.
Because if that’s all we are, we are easy to replace.
If instead we position SEO as the discipline that listens hardest to the market, translates that back into the business, and connects dots between product, marketing and customer reality, it becomes a lot harder to ignore us.
We are not just a marketing cost centre. We are not just another sales lever. We are, whether organisations realise it or not:
- the people who see demand shifts first,
- the people who see pain points in the raw,
- and the people who can turn all of that chaos into something a CEO can act on.
LLMs and AI search don’t change that. They just make it even more obvious that “traffic and attribution” by themselves are nowhere near the full story.
So maybe, next time you’re tempted to proudly present a visibility index and a ranking list as the proof of your value, add a different slide first:
“Here is what search and AI told us about our customers this month – and here’s what we should do about it.”
When we change that story, we stop being just another channel. And we finally start getting credit for the value we’ve been sitting on all along.