What is an AI visibility audit for a small business website?

76% of business owners now say it is essential their brand appears in ChatGPT answers. ChatGPT recommends 1.2% of local businesses. The gap between those two numbers is the entire reason the AI visibility audit exists.

If you have heard the term in a webinar or a competitor pitch and quietly wondered whether it is a real diagnostic or a buzzword, this post is for you. By the end you will know what an AI visibility audit measures, what the report looks like, what it costs, and the smallest first step you can take this week.

What an AI visibility audit actually is

An AI visibility audit is a diagnostic. It asks AI engines a set of real buyer questions about your category, then records what each engine answers. The output is a map of where your business is recommended, where it is mentioned but not recommended, and where it is missing entirely.

Picture a bookkeeper whose Google reviews are strong and whose local SEO is tidy. Someone opens ChatGPT and types "best bookkeeper for a small business in my area." The audit captures whether ChatGPT names that bookkeeper, names a competitor, or hedges with a generic answer. Then it does the same for Perplexity, Gemini, Claude, Grok, and DeepSeek.

That is the audit. Not a content order. Not a software subscription. A diagnostic that tells you whether the channel is working for your business right now.

What an AI visibility audit measures

A good audit measures four things. Most weak audits measure one.

The first is mentions. Does the AI name your business at all when asked about your category? A mention is the floor. Without it, nothing else matters.

The second is recommendations. A mention is "this business exists." A recommendation is "this is a good choice for your situation." The two are different. Strong audits report both.

The third is competitor coverage. AI does not rank in a vacuum. It picks from a small set of options, and the audit shows you which businesses it picks instead of yours, and why.

The fourth is the technical chain underneath. AI evaluation runs through five stages: discovery, comprehension, classification, trust, and selection. A breakdown at any stage means you do not get recommended, regardless of how strong you are at the others. The audit traces which stage is failing for your specific business.

Why small businesses need one in 2026

AI usage for local search grew from 6% in 2025 to 45% in 2026. AI search traffic overall grew 527% year over year. 49% of all ChatGPT activity is people asking for recommendations or advice. Half of the engine's traffic is the exact buyer behaviour you want to be visible inside.

Your channel mix changed in twelve months. Your marketing plan probably has not.

There is a second reason. Most small business owners assume their Google ranking transfers to AI visibility. It does not. AI engines use entity-based understanding, which means they see your business as a thing with attributes, not as a string of keywords matching a query. A business that ranks page one on Google for "wellness clinic Sydney" can be invisible to ChatGPT for the same question. The two systems evaluate differently.

If your last marketing investment was an SEO retainer, an AI visibility audit is the only way to know whether that work is also paying off in the channel buyers are increasingly using.

What the report looks like

A complete AI visibility audit report has six sections.

The first is the prompt panel. The exact buyer questions the audit ran across the engines. You should be able to read these and recognise them as questions a customer would ask.

The second is the engine breakdown. For each engine, a one-line verdict: recommended, mentioned, or missing. ChatGPT, Perplexity, Gemini at minimum.

The third is the competitor map. Which businesses got recommended in the answers where yours did not, and what the AI said about them.

The fourth is the technical findings. Crawl access, structured data, content clarity, internal linking. Plain language, not developer jargon.

The fifth is the trust profile. Reviews, third-party mentions, directory consistency. Business information should match across the open web. Eight in ten businesses have inconsistencies that quietly reduce AI confidence.

The sixth is the prioritised action list. Not a list of one hundred fixes. The three or four moves that move the needle, in order.

If a report you receive does not include all six, it is not an audit. It is a marketing pitch in audit clothing.

What an audit costs

Free tools exist. HubSpot's AEO Grader gives a one-prompt snapshot at no cost. Otterly starts at $29 a month for ongoing tracking. SE Ranking adds AI tracking to its standard SEO suite at around $65 a month. Managed services run higher. Snezzi sits at $999 a month for fully managed work.

For a small business that wants a one-off diagnostic with a usable report, expect to pay between $200 and $700 depending on how many engines, how many prompts, and how much remediation guidance is included. Ongoing tracking adds a monthly subscription on top.

The honest framing: a one-off audit answers "where do I stand right now." A subscription answers "is the work I am doing actually moving the score." You usually need the first before the second is worth paying for.

The five-stage AI selection chain

This is the part most audits skip and most owners need.

When someone asks an AI engine a question about your category, the answer passes through five stages.

Discovery. Can the engine even reach your site? About one in five small business sites have a crawl access problem that locks AI agents out before content is read. Robots rules, host settings, CDN configurations. Fixable. Often invisible without an audit.

Comprehension. Once the engine reads your site, can it understand what your business actually does? This is where structured data, clear headings, and plain-language service descriptions matter. AI prefers a two-thousand-word guide on one topic over ten two-hundred-word posts on related topics.

Classification. Does the engine put your business in the right category? A wellness clinic mis-classified as "spa" misses every "wellness clinic" recommendation prompt. The audit catches this.

Trust. Does external evidence corroborate what your site says? Reviews, directory consistency, third-party mentions. AI weighs evidence the open web exposes, not what your homepage claims.

Selection. Among classified, trusted, comprehensible options, why this business and not another? This is the layer where the competitor map matters.

Audit the chain, not the surface.

Common reasons businesses fail

Five patterns turn up in nearly every audit.

The first is flooding the site with thin AI-generated content. Volume hurts. Depth wins.

The second is assuming Google ranking equals AI visibility. Different system, different signals.

The third is measuring one engine and assuming full coverage. A brand can be number one in Gemini, missing in ChatGPT, mis-categorised in Perplexity in the same week.

The fourth is inconsistent business information across the web. Address, phone, services, hours. AI doubt rises with every mismatch.

The fifth is missing original expertise. Generic content is everywhere. Your case studies, your local knowledge, your actual proprietary data are what AI cannot find anywhere else, which is exactly why it cites you when it can.

What to do this week

Three steps. None of them require a credit card.

One. Open ChatGPT, Perplexity, and Gemini. Type the question your best customer would type to find a business like yours. Read the answers. Are you named? Are competitors named? Is the answer hedged?

Two. Note the gap. If you are not in any of the three answers, the audit's job is to tell you which of the five stages is breaking down for your business specifically.

Three. Run the free scan. Get Recommended runs three queries across ChatGPT, Perplexity, and Gemini at no cost (you enter one, the backend generates two more, for nine results total) and shows you whether your business is recommended, mentioned, or missing. The full report runs ten queries across six engines for sixty results, plus the technical and competitor layers.

Forward motion

The audit is not the work. The audit is the map. The work is the three or four moves it points to.

If you have a website, customers, and a category buyers ask AI about, an audit is no longer a luxury. It is the cheapest way to find out whether the channel that grew to 45% adoption this year is working for your business or working against it.

Run the free scan. See where you actually stand. Decide from there.