December 21

What Actually Sells in SEO Now (and What Quietly Stopped Making Money)

0  comments

SEO is in the middle of an effectiveness reckoning.

Not because optimisation stopped working. Not because Google is broken. And not because AI arrived and blew the whole thing up overnight.

SEO is struggling because the certainty that once allowed it to be sold with confidence has collapsed.

For a long time, the industry enjoyed a forgiving environment.

You could publish informational content, grow traffic, show graphs moving up and to the right, and allow the commercial story to remain comfortably vague. Attribution didn’t have to be perfect. The noise covered a lot of sins.

AI has stripped that noise away.

Informational demand is now being absorbed upstream by interfaces that answer questions directly.

Click paths are fragmenting. Fewer visits are reaching websites, and when they do, they arrive later, closer to decision, with far less patience for irrelevance.

This hasn’t killed SEO. It has removed its margin for bullshit.

And that’s why so many teams are suddenly struggling to justify budgets.

Not because organic search stopped generating revenue, but because the work being sold under the banner of “SEO” no longer maps clearly enough to how money is actually made.

This piece is about fixing that.

Not philosophically. Commercially.

The real problem isn’t AI. It’s commercial clarity.

One of the more revealing side effects of the last 18 months has been how often SEO conversations now collapse into defensiveness.

People talk down generative systems. They argue that GEO is “just SEO rebranded”. They insist nothing fundamental needs to change, even while admitting that search behaviour clearly is.

That contradiction matters.

If search is changing, and your product does not, the market will resolve the tension for you. And it will not be kind.

What’s really happening is simpler and more uncomfortable: many SEO propositions were never clearly commercial in the first place.

They relied on:

  • Informational traffic as a proxy for growth

  • Brand demand they didn’t create but happily claimed

  • Attribution fog that no CFO ever truly believed

When AI started absorbing informational queries, that scaffolding disappeared.

And now, I’m seeing more of those same SEO ‘gurus and consultants start writing and posting about budget cuts and ROI.

Those budgets are being cut because their work can’t be defended in plain commercial English.

This is one of my key gripes with the ‘it’s just SEO’ crowd.

Consultants and gurus had the opportunity to lead ‘GEO’ and create a new approach that got CMOs excited.

Instead, they started an industry civil war that they are now backtracking.

Of note, I love seeing the agencies of the ‘just SEO crowd’ now offer GEO/AI SEO services.

But fear not, Google has provided a life raft for SEO. And one I didn’t see coming.

Intelligent routing: the economic explanation everyone skips

To understand what actually changed, you have to stop thinking in terms of platforms and start thinking in terms of interfaces.

This is where Google itself has been unusually clear.

Nick Fox, Google’s SVP of Knowledge and Information, has spoken openly about the fact that there is no longer a single search experience. Different problems require different interfaces, and Google now routes queries accordingly.

This is what it calls intelligent routing.

In practice, it looks like this:

Short, urgent, high-intent queries These still behave like classic search.

  • “Emergency plumber near me”

  • “Buy running shoes size 10”

  • “iPhone 15 price”

Speed matters. Lists of links work. Ads monetise efficiently. Organic rankings still intercept demand directly. This behaviour is alive and well.

Longer, messier, exploratory queries These break the old interface.

  • “Which CRM is best for a small B2B team?”

  • “Is switching from X to Y worth it?”

  • “Compare iPhone vs Pixel for photography”

Here, ten blue links are a terrible experience. Users want synthesis, comparison, and recommendation. So these queries get routed into AI-led systems designed to answer, not retrieve.

This isn’t Google being generous. It’s Google being pragmatic.

And it explains almost everything that follows.

Why this forces SEO and GEO apart by function

Once you understand intelligent routing, the SEO vs GEO argument stops being semantic and becomes operational.

SEO increasingly lives at the bottom of the funnel.

Its job is to:

  • Capture buyer-intent demand

  • Place products and services where money is already on the table

  • Compete directly with paid search

  • Reduce ongoing acquisition cost

GEO lives in the decision-making layer.

Its job is to:

  • Influence how brands are described, compared, and recommended

  • Increase the probability of being chosen when systems synthesise answers

  • Load the information environment in your favour

Same market. Different interface. Different economics.

Trying to pretend these are the same thing is how spend gets diluted and results get muddy.

What SEO still sells — very clearly

Despite the panic, SEO has arguably never been clearer in what it’s good for.

1. Buyer-intent SEO (still the core revenue driver)

Buyer-intent SEO remains one of the most efficient demand-capture mechanisms in marketing.

People searching for commercial terms are telling you exactly where money matters. Brands reveal this every day through how much they’re willing to bid in paid search. Ranking organically for those queries isn’t clever — it’s financially sensible.

This work is unglamorous:

  • Product and service page optimisation

  • Clear propositions

  • Tight information efficiency

  • Internal linking that signals commercial importance

But it still maps cleanly to revenue, which is why it survives scrutiny.

If your SEO does not demonstrably support buyer-intent pages, you don’t have an SEO problem. You have a positioning problem.

2. Local SEO (largely unchanged)

Local search behaviour hasn’t fundamentally shifted.

People still search for:

  • Services near them

  • Opening hours

  • Reviews

  • Availability

Maps, listings, reviews, and proximity signals continue to do the heavy lifting. AI may summarise, but local intent remains immediate and action-oriented.

Local SEO still sells because the behaviour hasn’t meaningfully changed.

3. E-commerce SEO → agentic SEO

This is where the biggest transformation is coming.

Classic category-page optimisation won’t disappear, but e-commerce increasingly shifts toward availability, feeds, and agent-mediated shopping.

Shopping feeds, structured inventory data, pricing, fulfilment signals — these become the raw materials both search engines and AI agents rely on to make decisions.

SEOs who understand:

  • Product data

  • Merchant feeds

  • Inventory logic

  • Platform integrations

…will find plenty of demand.

This isn’t content-led SEO. It’s infrastructure-led SEO.

4. Technical SEO (still boring, still essential)

Websites still break.

They still suffer from crawl inefficiencies, indexation issues, performance bottlenecks, and security failures. AI doesn’t change that. It amplifies the cost of getting it wrong.

Technical SEO remains hygiene, but hygiene matters more when margins are tighter.

It sells because it protects access — for users and machines alike.

What informational content is actually for now

Informational content isn’t dead. The myth attached to it is.

Creating informational content purely to capture traffic is no longer defensible for most businesses. AI systems now extract that value before users ever arrive.

But content still matters structurally.

Its job is now to:

  • Support commercial pages via internal linking

  • Reinforce credibility

  • Enable PR and coverage

  • Build fame, feeling, and fluency

  • Arm sales teams

  • Nudge buyers already considering a purchase

Content is no longer a traffic engine by default. It’s infrastructure.

And infrastructure compounds.

What GEO actually sells (and why it’s marketing)

Generative engine optimisation is not SEO.

GEO is the practice of increasing the likelihood that AI systems recommend your brand in a buying situation.

Not rank it. Not “surface” it. Recommend it.

There is no persistent position to optimise for. There is only probability.

AI systems don’t see real-world performance. They don’t know your revenue or retention. They infer quality from what exists in the information environment around your brand.

Mentions. Repetition. Corroboration. Third-party validation.

This is availability logic applied to machines — the same principle long established in marketing science.

Which is why GEO sits much closer to branding, PR, and copywriting than traditional SEO mechanics.

Why copywriting becomes central again

When an AI system recommends something, it has to justify that recommendation.

It does this by referencing attributes:

  • Who it’s for

  • What it’s good at

  • What trade-offs exist

  • Why one option fits better than another

If your copy is vague, interchangeable, or hedged, the system has nothing to work with.

Clear positioning increases confidence. Confidence increases recommendation probability.

This work may not improve rankings. That’s fine. Ranking is no longer the objective.

Being chosen is.

Why PR becomes influence infrastructure

Digital PR matters because it places brands into authoritative contexts.

From an SEO perspective, this still supports buyer-intent rankings. From a GEO perspective, it increases corroboration across trusted sources.

Very few activities do both.

That’s why PR isn’t a bolt-on anymore. It’s influence infrastructure — and infrastructure is what compounds when interfaces keep changing.

The definitive list: what actually sells now

If you strip everything back, this is what survives commercial scrutiny:

  • Buyer-intent SEO (product and service placement)

  • Links

  • Local SEO

  • E-commerce SEO evolving into agentic SEO

  • Technical SEO as access protection

  • Copywriting that articulates real differentiation

  • Digital PR that earns corroboration

  • Content marketing built for fame, feeling, and fluency — not traffic

Everything else needs to justify itself against these.

The choice SEOs now face

SEO as a single, fuzzy discipline is fragmenting.

That’s not a threat. It’s a clarification.

You can:

  • Focus on infrastructure

  • Focus on placement

  • Focus on influence

  • Focus on content as a brand asset

What you can’t do anymore is rely on informational noise to make the numbers look plausible.

The market has moved on. Don’t die on the citation and fan out hill.

No matter how many blog posts you want to sell…that ship isn’t sailing for long.

The calm conclusion

Organic search is alive and well. Money is still being made.

But it’s being made in fewer, clearer places — closer to purchase, closer to decision, closer to influence.

AI didn’t break SEO. It removed the excuses.

If you understand how demand is captured and how choice is influenced, there is plenty to sell.

If you don’t, the market will teach you very quickly.

And it will not use gentle language.

Andrew Holland


Originally published on LinkedIn on 21 December 2025.


Tags

LinkedIn archive


>