What is AI search optimisation
AI search optimisation is the work of making your business easy for AI models to cite. When a customer asks ChatGPT, Perplexity, Claude or Google AI Overviews for a recommendation, the model composes a short answer that names two or three specific businesses. AI search optimisation is how you become one of the named businesses.
There are two sub-disciplines you will see branded separately. They overlap heavily and you can mostly ignore the distinction.
- GEO, Generative Engine Optimisation. Optimising content so generative models will quote it.
- AEO, Answer Engine Optimisation. Optimising content so answer engines give it as the answer.
Both come down to the same activities. Plain-English content. Structured data. Entity clarity. Consistent data across the open web. Everything else is terminology. If you want to check where your business stands today, the free AI Visibility Checker takes under 60 seconds.
GEO, AEO and AI visibility: what the terms actually mean
The acronyms are new, the underlying work is not. Here is how the field breaks down in 2026, in plain English.
GEO is about being quotable
Generative Engine Optimisation focuses on the raw content a model reads. Short paragraphs. Clear headings. Plain-English sentences. Information a model can quote without editing. A 600-word wall of trade jargon is not quotable. A tight FAQ with six real customer questions is.
AEO is about being the direct answer
Answer Engine Optimisation focuses on the question-to-answer mapping. For every question a customer asks, there should be a single clear answer on your site, ideally inside a FAQPage schema block. This is what lets a model say "according to this business" rather than "according to various sources".
AI visibility is the outcome
AI visibility is the measurable result. It is how often your business is named, which queries you are named for, and which models are citing you. For a small business, this is the only number that matters. GEO and AEO are the inputs. Visibility is the output.
Why small businesses have the advantage
Counter-intuitive, but true in 2026. Small, local, niche businesses have a structural advantage in AI search. There are three reasons.
First, AI models favour specificity. "Independent accountant in Leeds" is a more useful answer than "Big 4 accountancy firm" for a small business owner looking for a tax return. Specificity is easier for a small business to claim than for a national chain.
Second, entity signals matter more than content volume. A small business that nails its Google Business Profile, FAQ schema, and areaServed can outperform a content mill publishing ten articles a week, because the small business is a clean entity and the content mill is a noisy one.
Third, AI models are hungry for trust signals that a local business can credibly produce. Real reviews. A named owner. A consistent address. These are hard for faceless content farms to fake and easy for a real business to supply.
The four fixes that do 80 percent of the work
If you only do four things, do these. Each one is under an hour. Combined, they move more AI citations than the other 20 fixes on a full audit. For a walkthrough of every signal, see our guide to showing up in ChatGPT.
1. Homepage hero, first 40 words
Plain English. Location. Three services. No marketing adjectives. Write how a customer would describe what you do when recommending you to a friend.
2. Six FAQs with JSON-LD schema
Real customer questions. Wrapped in FAQPage schema. Pasted into the head of your services page. This is the single highest-value schema block for a small business. For the technical walk-through, see our schema markup guide.
3. Google Business Profile, finished properly
Primary category as specific as possible. Every secondary slot used. Description written to the full 750 character budget. Opening hours correct. Service area named. Photos uploaded. If you have not touched it in a year, it is out of date.
4. One service per page
A single homepage listing every service is weaker than one page per service. Split emergency call-outs, boiler service, bathroom installs onto their own URLs. Each page gets its own Service schema. This lets AI models answer narrower queries with you as the specific recommendation.
Start with the free check.
See whether ChatGPT, Perplexity and Claude are citing you today, before you spend a weekend on fixes.
The fixes that look useful but are not
A lot of AI search advice on the internet in 2026 is wrong or outdated. Here is what to ignore.
- Keyword density. AI models do not count keyword frequency. They weigh semantic match. Write for humans, the model will follow.
- AI-generated content at volume. Models are now good at detecting low-quality AI content and downweighting the sources. A content farm is a net negative in 2026.
- Generic schema from a plugin. Out-of-the-box Organization schema with default fields is close to useless. Specific, complete, well-typed schema is what moves the needle.
- Backlink campaigns. Links still matter for classic SEO. They are a weaker signal for AI citations than consistency and structured data. Do not redirect a small business budget to link building.
The entity signal, and why it matters more than content volume
AI models think in entities. A business, in model terms, is a cluster of facts that all point at the same real-world thing. Name. Address. Category. Services. Reviews. Opening hours. When these facts line up across every source the model can see, your business becomes a clean entity. When they conflict, you are a noisy entity.
Clean entities get cited. Noisy entities get skipped.
Three places where small businesses break entity clarity without meaning to.
- Trading name versus legal name. The website uses "Smith Plumbing", the Google Business Profile uses "J Smith Plumbing Ltd", and Yell has "Smith and Sons Plumbing". Pick one and use it everywhere.
- Phone number variations. One listing has the mobile, another has the landline, a third has a call-tracking number. AI models treat these as separate entities. Use one primary number.
- Address formatting. "12 High Street", "12, High St", "High Street 12". Same place, three strings. Consistent formatting is a cheap fix that prevents the model from hesitating.
Fix these three and your entity signal strengthens without writing a word of new content. It is often the single highest-leverage hour of work available to a small business.
A related trap. Businesses that use a separate booking platform or call tracking often let the third-party listing drift out of sync with the main site. If your GBP phone number differs from the one in your LocalBusiness schema, both are weaker than if they matched. Use the same string everywhere, even if it costs you one call-tracking feature.
AI visibility for specific industries
The four fixes above are universal. The specifics differ by trade. Which Google Business Profile categories apply. Which FAQs customers actually ask. Which LocalBusiness sub-type to use. We maintain industry-specific pages for each of the common small business categories.
How long it takes and what to expect
Implementation is a weekend. Most individual fixes are under an hour. The longer tail is waiting for AI models to re-crawl and re-index your site. Expect one to three weeks for ChatGPT to pick up a homepage rewrite. Perplexity and Google AI Overviews tend to be faster. Claude is slower.
Do the citation check once a week for the first month, then monthly. Track which queries you win and which you lose. That log is the input to the next round of improvements.
The 30-day plan for a small business
A realistic plan for a business owner with a few hours a week. Four weeks, four focus areas, measurable at the end.
Week 1, baseline and the homepage
Run the free checker. Log which queries your business is currently cited for, and which it is not. Rewrite the first 40 words of your homepage in plain English. Add a specific location. Ship it the same day.
Week 2, FAQ schema and services
Write six real FAQs. Wrap them in FAQPage JSON-LD. Paste into the services page. Split any multi-service homepage into individual service pages with their own Service schema blocks. Validate with the Rich Results Test.
Week 3, Google Business Profile and reviews
Rewrite the GBP description to 750 characters. Add every applicable secondary category. Upload fresh photos. Email the last 20 happy clients with a review prompt. Real prompts, real reviews, no fakes.
Week 4, measure and iterate
Re-run the ten priority queries through ChatGPT, Perplexity and Claude. Log any citation changes. Identify two more queries to target for the next round. Pick your next priority service page to rewrite. Repeat the cycle.
At the end of 30 days, most small businesses see citation changes in at least two tools. The effect compounds over months. Keep the cycle running.
Measuring AI visibility without a dashboard
You do not need a paid tool to track AI visibility in 2026. A spreadsheet is enough. Five columns.
- Date of check
- Query pasted
- Tool used, ChatGPT, Perplexity, Claude, Google AI Overviews
- Were you cited, yes or no
- Who else was cited
Fifteen minutes a week, ten priority queries, four tools. That gives you 40 data points per week. Over a month, that is enough to see trends. Over a quarter, it is enough to prove the work is paying.
The same spreadsheet tells you who your real competitors are in AI search. Often they are not the competitors you thought. Small independents with good FAQ schema frequently outrank bigger players who have not done the work. That intel is worth its own line in your monthly plan.
When to DIY, when to pay
DIY is fine for a confident business owner with a weekend free. The $15 workbook shortens the learning curve and gives you paste-ready templates. Most small businesses can ship the full checklist from the workbook alone.
Pay for the $197 audit if you want a priority-ranked list of exactly what to ship first, with the specific copy and schema blocks ready to paste into your CMS. Turnaround is 48 hours.
Pay for done-for-you implementation only if your time is genuinely worth more than $80 an hour and you have no developer available.
Pick your path.
Free checker, $15 workbook, or $197 audit. Most small businesses start with the free check.