AI Overview Optimisation Is More Like Horse Whispering Than SEO

AI Overview Optimisation is a terrible name.

Not because it is wrong exactly, but because it makes the whole thing sound far more mechanical than it is. It sounds like something you can force.

Add some FAQs.
Mention the keyword.
Write “what is” at the top of the page.
Wait for Google to reward you.

Lovely. Except, in our experience, that is not really how it works.

AI Overview Optimisation is not about grabbing Google by the reins and dragging it towards your website. It is much closer to horse whispering.

You watch what is already happening.
You understand where the movement is.
You look for the little signals.
You remove the friction.
You gently guide the page in the direction it already wants to go.

That is the bit a lot of people miss. AEO, GEO, AI Overview Optimisation, Answer Engine Optimisation, or whatever we all end up calling it next week, is not just a shiny new version of SEO.

It is search strategy, content strategy, conversion strategy and a healthy amount of common sense, all rolled together. Except, it leaves behind the “Black Magic” of SEO. In fact, Generative Engine Optimisation is probably the purest form of SEO. It is about creating a name for yourself and using content as the truest format to show you are an expert and have authority to speak on a topic.

It is SEO strategy, content strategy, conversion strategy and a healthy amount of common sense, all rolled together.

The mistake people make with AI Overview Optimisation

The mistake is assuming that AI Overview Optimisation starts with new content.

Sometimes it does.

But often, the best opportunities are already sitting inside your website.

They are the blog posts getting regular traffic.
The service pages ranking, but not quite converting.
The old guides that Google still seems to like.
The second-page rankings that are close enough to smell page one.
The articles that answer the right question, but not clearly enough.
The pages that have authority, but no proper next step.

This is also why Google’s own guidance is worth paying attention to. In its official advice on optimising for generative AI features in Search, Google says that SEO is still relevant because generative AI features are rooted in its core Search ranking and quality systems.

Which is Google’s very polite way of saying: please do not abandon the basics and start doing weird things. That is why this is not just about creating more content. It is about finding the pages and topics where Google already seems to trust you, then making those pages clearer, more useful and more commercially joined up.

Because if Google is already showing your page impressions, rankings or traffic, that is a signal.

It might not be a standing ovation.

But it is at least a nod from across the room.

What AI Overview Optimisation actually looks like

In practical terms, AI Overview Optimisation usually starts with a few fairly boring marketing basics.

Not magic prompts.
Not secret hacks.
Not “the ultimate AI ranking formula”.

Just basics.

We look at:

• Pages already getting regular organic traffic
• Queries already appearing in Google Search Console
• Rankings sitting on page two or low page one
• Topics the website is already known for
• Search results that trigger AI Overviews
• Article formats that are already performing, such as “best”, “free”, “how to” and “vs” content
• Pages with commercial value, not just search volume
• The journey from information page to conversion page

That last point is important.

Because appearing in an AI Overview is nice. But if the page does not move the user anywhere useful, it is just visibility with nowhere to go and we have all seen enough marketing activity that creates visibility with nowhere to go.

Usually right before someone asks why the leads have not improved.

This is where content creation and SEO cannot be treated as two separate jobs. The article has to be useful enough to rank, clear enough to be understood, and commercially connected enough to support the next step.

AI Overviews are changing the shape of search

Google describes AI Overviews as a snapshot of key information with links that help users explore more on the web. That sounds neat and tidy.

The reality is messier.

A recent large-scale study of Google AI Overviews looked at more than 55,000 trending queries and found that AI Overviews appeared for 13.7% of queries overall, but rose to 64.7% for question-form queries. That is the bit marketers should pay attention to.

Questions matter.

The search landscape is not just moving towards keywords. It is moving towards answers, summaries and selection. That does not mean every business needs 900 FAQ pages. Please do not do that.

But it does mean that the structure of your content matters more than ever.

If the user is asking a clear question and your page gives a better, clearer, more trusted answer than the alternatives, you are much more likely to be in the conversation.

That is the principle behind our blog on how to optimise for the AI Overview. The aim is not to game the system. The aim is to answer the right questions, in the right structure, with the right authority behind them.

Route one: improve what is already working

The first route is taking content that already has traction and making it better.

That might mean a blog post that gets regular impressions but has a weak structure. It might be a page that answers a valuable question, but hides the answer three paragraphs below a vague introduction.

It might be a guide that ranks well, but has no internal links, no embedded CTA and no clear relationship to the service you actually want to sell.

This is where the “horse whispering” bit comes in.

You are not trying to turn a page about garden furniture into a page about AI apprenticeships. You are listening to what Google is already telling you.

If a page is getting impressions for useful questions, that tells you there is interest.
If a page is sitting on page two, that tells you there is relevance.
If a page is being found through long-tail queries, that tells you there is intent.
If a page is getting traffic but no enquiries, that tells you the journey is broken.

The optimisation work is then about making the page the best possible version of itself.

Clearer headings.
Better answers.
Stronger internal links.
Useful FAQs.
More evidence.
More expert commentary.
A more obvious next step.

Not more words for the sake of it.

Better words in a better order.

That is also why we usually start with an Explore audit, rather than jumping straight into “let’s write 12 blogs”. The audit finds the movement first. Then the strategy follows.

Case study one: the deeper charity page that became a donation asset

We saw this clearly in a whitelabel AEO project for a charity-sector website.

The opportunity was not a brand-new landing page built from scratch. The opportunity was a deeper resource page that already had a reason to exist.

The work focused on making that page more answer-led, more useful, more structured and more connected to the donation journey. The result was not just more traffic.

It was better traffic.

The page accounted for only a small proportion of views across the tracked topic area, but delivered the majority of key donation events and donation revenue.

That is the bit that matters.

This was not a case of “we ranked for a keyword, everyone clapped, then nothing happened.” The page answered the right question at the right moment, then gave the user a meaningful next step.

That is what AI Overview Optimisation should be aiming for.

Not just visibility.

Useful visibility.

Conversion rate is where this gets interesting

This is where the conversation needs to move away from traffic alone.

Because traffic is not the prize.

Leads, enquiries, sales, donations, bookings and opportunities are the prize.

The Unbounce Conversion Benchmark Report found a median conversion rate of 6.6% across all industries, based on 464 million pageviews, 41,000 landing pages and 57 million conversions.

That is a useful benchmark, but context matters. A B2B service page, an ecommerce product page, a charity donation journey and a course enquiry page are not doing the same job.

First Page Sage’s conversion rate by channel data gives a more B2B-specific comparison, putting B2B organic search conversion at 2.6%, PPC at 1.5%, organic social at 1.7% and paid social at 0.9%.

Meanwhile, Ruler Analytics’ 2026 conversion rate benchmark data shows just how much conversion varies by sector, with 2025 organic search conversion rates listed at 2.3% for B2B services, 2.5% for B2B tech and 5.0% for professional services.

The numbers will always vary depending on how a business tracks conversions. But the direction of travel is clear.

For B2B SMEs, organic visibility still matters because it tends to sit closer to intent than most interruption-based channels.

And AI search traffic, while often smaller in volume, can be even more interesting because the user may already have done some of the thinking before they click.

That is something we covered in more detail in our blog on AI traffic vs SEO vs social for B2B SMEs.

Route two: find the topic you are already trusted for, then build sideways

The second route is slightly different.

Instead of starting with one page, you start with the topic.

This works especially well when a business wants to be known for something specific, but does not yet have the perfect page to capture that demand.

The question is not:

“What keyword do we want to rank for?”

The better question is:

“Where does existing authority overlap with a searched topic and a commercial objective?”

That overlap is where things get interesting.

Because you might want to rank for something huge and broad.

But if Google does not currently associate your website with that topic, it is going to be a slog.

A better approach is to look at what the website is already recognised for, then find related questions that move the business towards where it wants to be.

That is exactly what we looked at with The Coders Guild.

Case study two: The Coders Guild and AI apprenticeships

We saw the same principle at work with The Coders Guild.

Rather than simply writing more content about AI, we looked at where the website already had traction, which pages were getting regular traffic, which rankings were sitting just outside page one, and which question-led searches matched the organisation’s commercial objective: building visibility around AI training, AI apprenticeships and AI skills for SMEs.

One opportunity came from a simple but valuable question:

“Where can I learn AI and automation for free?”

On the surface, it is an early-stage search.

But for a training provider, that is exactly where the journey often begins. By creating an information page designed to answer the question properly, embedding CTAs, connecting it to a clearer course page and supporting it with webinars and related blogs, the activity turned curiosity into movement.

The result was a 30% increase in organic search traffic quarter-on-quarter, a 56% increase in form submissions, a 135% increase in ChatGPT referrals, a 566% increase in AI form fills, four converting leads from AI, and a 2% traffic-to-lead conversion rate.

Something to Share…

Why a 2% traffic-to-lead conversion rate matters

A 2% traffic-to-lead conversion rate might not sound dramatic if you are used to reading conversion rate posts that promise the moon on a stick.

But in B2B, it is meaningful. Especially when the traffic is coming from new AI-led sources.

If B2B organic search benchmarks sit around the 2% to 3% mark in many datasets, then a page attracting early-stage AI and organic visibility while still converting at 2% is doing its job.

It is not just getting readers. It is creating movement.

That is the part many businesses miss when they judge content too quickly.

A top-of-funnel information page is not supposed to convert like a branded PPC landing page.

It has a different job.

It should attract the right audience, answer the right question, build enough trust to continue the journey, and then move some of that audience towards a more commercial next step.

If it does that, it is not “just a blog”. It is part of the sales journey.

AI traffic is not the same as SEO traffic

One of the most interesting things we are seeing is that AI-driven traffic often behaves differently.

It is usually lower volume but it can be much higher intent. That makes sense.

If someone arrives via a traditional organic search, they may still be exploring the basics.

If someone arrives via ChatGPT, Gemini, Perplexity or an AI Overview-style journey, they may already have done a lot of the thinking before they click.

They may have compared options.
They may have asked follow-up questions.
They may have narrowed the problem.
They may already know the type of solution they need.

By the time they reach the website, they are not always starting from scratch.

They are validating.

This is important for B2B because buyers are often doing more of the journey before they speak to anyone. Forrester has reported that 41% of buyers have a single vendor in mind when they first begin the purchase process, which is terrifying if you are not already in that buyer’s head before they make contact.

That is why AI search matters.

It is not just another traffic source. It is part of the research layer where buyers form opinions, compare options and decide who deserves to make the shortlist.

For B2B SMEs, this is the important bit.

The future of search is not just about who gets the most traffic. It is about who gets selected when the buyer is forming their shortlist.

You still need SEO

This is where the conversation can get a bit silly.

Every time something new appears in marketing, someone declares the old thing dead.

SEO is dead.
Blogs are dead.
Social is dead.
Email is dead.
Probably fax is making a comeback.

But AI Overview Optimisation does not replace SEO.

It sits on top of it.

You still need a technically sound website.
You still need useful content.
You still need topical authority.
You still need internal links.
You still need pages that load properly.
You still need to understand what people are searching for.
You still need to convert the traffic when it arrives.

The difference is that the search result is changing.

Google is no longer just listing pages.

It is summarising, comparing and selecting.

That means your content needs to be easier to understand, easier to trust and easier to use as part of a wider answer.

This is why our approach to SEO and strategy has always been broader than “rankings only”. Rankings matter. Of course they do. But they matter because of what they help the business achieve.

You also need a proper journey

This is the other bit people miss.

There is no point capturing an AI Overview if the user lands on a page that gives them nowhere to go.

A good AEO page should not feel like a hard-sell landing page.

But it should have a job.

That job might be:

• Read the related guide
• View the course
• Download the resource
• Book a webinar
• Enquire about the service
• Compare options
• Speak to the team
• Join the mailing list

The CTA does not have to shout.

It just has to be there when the reader is ready.

In The Coders Guild example, the information page worked because it did not sit in isolation. It connected to a course page and supporting activity.

That is where a lot of SME websites fall down.

They have useful content.
They have decent service pages.
They have forms.
They have webinars.
They have case studies.

But none of it is joined together properly.

AEO rewards clarity.

Users do too.

This is also where multi-channel marketing becomes important. The page might win the search. But the webinar, email, social post, paid remarketing and sales follow-up might be what turns that interest into a lead.