AI Visibility - Is your business invisible to AI search?
AI visibility is being named when buyers ask ChatGPT, Perplexity or Google’s AI who to use. Assist IQ audits where your business stands across the engines, scores it out of 100, then builds the review signals, third-party mentions and on-site answers the evidence says actually move it. Measured monthly and reported honestly, for small businesses across Scotland and the UK.
AEO and GEO, plainly - The engines answer. The question is whether you're in it.
When a buyer asks an assistant “who should I use for X near me”, the answer is synthesised from what the engines can find and verify: your site, your reviews, directories, and what third parties say about you. Three things decide the outcome.
- Present. You have to exist in the places engines actually read: an indexable site, complete business profiles, the directories and platforms your sector lives on. Absence anywhere is absence in the answer.
- Parseable. A machine has to be able to lift a clean answer from your content: who you are, what you do, where, for whom, with clear structure and no JavaScript required to read it.
- Corroborated. Engines trust what other people say about you more than what you say about yourself. Reviews, mentions and third-party listings carry most of the weight. Classic SEO gets you eligible. Corroboration gets you chosen.
The honest bit - What we will not sell you.
Plenty of SEO shops have rebadged as AI-visibility experts this year. The fastest way to test one is to ask what they refuse to sell. Here is our list.
- Guaranteed AI rankings. Ask the same assistant the same question twice and you can get different names. Nobody controls that, including us. We measure how often you are mentioned across repeated runs, and we report that number honestly, up or down.
- llms.txt as a magic file. We ship one because it takes five minutes and costs nothing. We will never tell you it drives citations, because there is no good evidence it does.
- Schema as a ranking trick. Structured data helps machines read facts about your business, and we do it properly on every page. It is hygiene, not a lever. Anyone selling it as the main event is selling the easy part.
The method - A score out of 100, then a plan in evidence order.
Every engagement starts with the same six-pillar audit. The weighting is the honest part: reviews and citation coverage carry 45 of the 100 points. The technical work most agencies lead with carries 10.
| Pillar | What we assess | Points |
|---|---|---|
| A. Answer-surface baseline | Your key buyer questions, run across ChatGPT, Perplexity, Google AI Overviews and Copilot. Who gets named, what gets cited, and whether anything said about you is wrong. | 15 |
| B. Citation-source coverage | Which sources the engines actually retrieved for those answers, and whether you are present in them. The gap between the two lists is the plan. | 20 |
| C. Review and proof signals | Count, recency, spread and response rate, against your competitors. Proof trapped where engines cannot see it is the most common problem we find. | 25 |
| D. Entity and machine-readability | Business profile completeness, Bing and Google indexation, AI crawler access, consistent name-address-phone everywhere, correct structured data. | 20 |
| E. On-site answerability | Whether your website answers the questions buyers actually ask assistants: answer-first pages, comparisons, FAQs a machine can lift. | 10 |
| F. Competitive reverse-engineering | For every competitor named ahead of you: why them? Each answer becomes an action. | 10 |
You get the scored baseline, dated, plus the gap list in evidence-ranked order and the on-site build spec. Then the work: the review engine, the citation build-out, the site fixes, and a monthly re-run of the same battery so you can see what moved.
We run this exact method on ourselves, monthly.
The same six-pillar audit, the same query battery, pointed at Assist IQ, against the questions our own prospects ask assistants. It keeps us honest: if we cannot move our own number, we have no business charging you to move yours.
This site is built the way we tell clients to build: answer first, readable without JavaScript, structured data on every page. The site you are reading is the sample of work.
Questions we get asked about AI visibility.
The honest-uncertainty answers. If an agency answers these differently, ask them for their evidence.
How is AI visibility different from SEO?
Classic SEO targets the list of blue links. AI visibility targets the answer engines give before anyone clicks. The foundations overlap: fast, readable pages, structured content, credible mentions. But the weightings differ: review signals and third-party corroboration matter more for being cited, and Bing indexation matters because it feeds several assistants. Good SEO makes you eligible. It does not make you chosen.
Do I need a new website for this to work?
Not always. Answer-first copy, FAQ blocks, comparison content and clean structure can usually be added to an existing site. What is non-negotiable is that your content is readable without JavaScript, because many AI crawlers do not run it. Our audit checks that first. If you are rebuilding anyway, baking this in from day one costs nothing extra.
How long before we see results?
Review signals and profile work start compounding within the first weeks. Citation coverage builds over months. We run the same query battery at the start and again at 90 days, and that before-and-after is the deliverable. What we will not do is promise a specific engine will name you by a specific date. Nobody can honestly promise that.
Is this worth it for a small local business?
Usually yes, precisely because almost nobody local has done the work. Most UK local markets have close to zero competition for AI citations, so the first mover in a vertical owns the position. The caveat: you need real proof behind you, reviews and customers worth naming. If that base is missing, build it first. We will tell you which side of the line you are on.
Tell us about your business
Not a discovery call. Not a pitch with a calendar link. The AI Ops Review is 45 minutes of questions about how your business actually runs. Then you get a written plan of what’s worth doing, what it costs, and what isn’t worth doing at all.
It’s £750. If we can’t find you at least 5 hours a week, it’s free. If you build with us inside 60 days, the £750 comes off the bill. So the only way it ever costs you money is if you take the plan and run it yourself. Fair enough.
Worst case, you lose 45 minutes and leave with straight answers.
No prep. No data handover. No long contract. You talk, we listen.