What is AI discoverability?

AI discoverability is how likely an AI assistant such as ChatGPT, Gemini, Claude, Copilot or Perplexity is to surface your business when someone asks a relevant question. It depends on five things: whether your website is indexed in the search indexes those assistants draw on, whether your business is identifiable as an entity with a name, location and services, whether your content is structured so answers can be extracted, whether independent sources corroborate that you exist, and whether you have a presence inside the assistants themselves.

How do I check if AI can find my business?

Ask ChatGPT, Gemini and Perplexity to recommend a business like yours in your city, phrased the way a customer would. Then search your exact business name in both Google and Bing. If your own site is not the first result in both, your indexing or entity signal is the problem, and that is where to start.

Is my business indexed in Bing?

Search your business name in Bing directly. If nothing appears, create a free Bing Webmaster Tools account, verify your site and submit a sitemap. This matters more than most businesses realise, because ChatGPT relies substantially on Bing indexed results when answering business questions.

How is AI discoverability different from SEO?

Traditional SEO optimises for ranked links. AI discoverability, sometimes called generative engine optimisation, optimises for being quoted inside a generated answer. It rewards explicit facts, self contained answers, structured data and corroboration across independent sources. A site can rank respectably in search and still never be cited by an assistant.

How long does it take to improve my score?

Indexing fixes can take effect within days. Entity and structure work takes a few days of focused effort. Corroboration through reviews, listings and mentions builds over months. Expect visible movement within weeks and meaningful presence over a quarter.

Do I need to hire someone to fix this?

The Layer 1 to Layer 3 items are genuinely doable yourself if you have access to your website and a few hours. Layer 4 is mostly persistence rather than skill. Layer 5 is where most people stall, not because the tools are hard but because connecting the assistants to a booking flow, and testing what they say, is fiddly work that never reaches the top of a busy week.

What score should I aim for?

Above 35 puts you ahead of most small service businesses in 2026. Above 45 is rare. Below 15 means marketing spend is being wasted, because attention arrives somewhere it cannot be captured.

AI Discoverability Checklist: Score Your Business Free

Most businesses have no idea whether AI assistants can find them, and the usual advice does not help, because it tells you what to do without telling you where you currently stand.

This is a self audit. Twenty five checks across five layers, scored out of 50. It takes about 30 minutes and requires no tools you have to pay for. At the end you get a band, an interpretation, and the specific next action for your band rather than a generic list.

It was built by Clause & Code, a UAE company that builds AI client systems for coaches, consultants and professional practices, from the audit we run before every build. The failures are remarkably consistent, which is what makes a checklist useful in the first place.

Scoring: each check is worth 2 points. Award the full 2 only if the answer is clearly yes. Half credit exists where stated. Be strict with yourself. A generous audit produces a comfortable score and no improvement.

Before you start: the 60 second reality check

Do this first, because it sets your expectations for everything below.

Open ChatGPT, Gemini and Perplexity. In each one, ask a question a real prospect would ask, phrased naturally. For example: "Who are good business coaches in Dubai for female founders?" or "Recommend an architecture studio in Sharjah for villa design."

Write down whether your business appears in any of the three. For most readers it will not appear in any. That is your baseline, and the checklist below explains why.

Layer 1: Indexing, 10 points

Nothing else matters until this layer passes. An assistant cannot quote a page it cannot reach.

Your site appears when you search your exact business name in Google. Search your full business name in quotes. If your own site is not the first result, your indexing or your entity signal is broken. 2 points.
Your site appears when you search your exact business name in Bing. Do the same search in Bing. This check is disproportionately important, because reporting indicates ChatGPT draws heavily on Bing indexed results for business queries. Most businesses have never checked this. 2 points.
You have a verified Google Search Console property with a submitted sitemap. Not just an account. Verified, sitemap submitted, and no coverage errors on your key pages. 2 points.
You have a verified Bing Webmaster Tools property with a submitted sitemap. The single most commonly skipped item on this entire checklist. 2 points.
Your page content appears in the raw HTML. Right click your homepage, choose view page source, and search for a distinctive sentence from your page. If you cannot find it, your content is being drawn in by JavaScript after load and some crawlers may never see it. 2 points.

Layer 1 score: ___ / 10

Layer 2: Entity clarity, 10 points

This layer determines whether a machine can state who you are without guessing.

Your legal or trading business name appears as text on your site. Not only in a logo image. As readable text. 2 points.
Your city and country appear as text on your homepage. Extremely common failure among service businesses that consider themselves location independent. If you serve a market, name it. 2 points.
You have Organization or ProfessionalService schema on your site. Test it by pasting your URL into Google's Rich Results Test. If nothing structured is detected, score zero. 2 points.
Your schema includes address, contact details and sameAs links to your verified profiles. Schema that exists but omits these does half the job. 1 point for schema present without them, 2 points for complete.
Your name, address and phone details are identical everywhere you appear online. Website, Google Business Profile, LinkedIn, directories, Instagram bio. Inconsistent details fragment your entity across sources and weaken all of them. Check three places before scoring. 2 points.

Layer 2 score: ___ / 10

Layer 3: Content structure, 10 points

This layer determines whether you are quotable once you are found.

Your page titles contain the words a buyer would actually type. Your service, your audience, and your location. A title containing only your brand name or a coined product term scores zero, because nobody searches for a phrase you invented. 2 points.
You have at least one FAQ section with real questions. Questions phrased as people ask them, not as marketing prompts. 2 points.
Your FAQ is marked up with FAQPage schema. Verify in the Rich Results Test. Unmarked FAQ content is worth roughly half as much. 2 points.
Your answers are self contained. Take any answer on your site and read it alone, with nothing above or below it. Does it still make complete sense? Assistants quote in isolation. Answers that depend on surrounding context get skipped. 2 points.
Your concrete facts are stated plainly in text. Prices or price ranges, timelines, what is included, what is not. Vagueness is the single biggest reason a page is never cited, because an assistant cannot quote a fact you did not state. 2 points.

Layer 3 score: ___ / 10

See our companion guide on building landing pages structured for AI extraction if this layer scores low.

Layer 4: Corroboration, 10 points

This is the layer almost everyone underestimates, and the one that most determines whether an assistant will actually recommend you rather than merely describe you.

You have a claimed and fully completed Google Business Profile. Claimed is not enough. Categories, services, hours, photos, description. 2 points.
You have at least ten genuine reviews across any platform. 1 point for one to nine, 2 points for ten or more. Most small service businesses have zero, not because clients were unhappy, but because nobody asked.
You appear in at least three credible directories relevant to your country or industry. Free zone or chamber listings, industry associations, B2B directories. 1 point for one or two, 2 points for three or more.
Your LinkedIn company page is complete and links to your site. A stub page with a logo and nothing else scores zero. 2 points.
At least one source you do not control mentions your business by name. A client site, a partner page, a podcast description, an article, a directory writeup. If every mention of your business was written by you, an assistant has exactly one source and treats it as a claim rather than evidence. 2 points.

Layer 4 score: ___ / 10

Layer 5: Presence inside the assistants, 10 points

Layers one to four determine whether you get found. This layer determines what happens next.

You have a published custom GPT trained on your expertise. Shareable, live, and trained on your actual methodology rather than your marketing copy. 2 points.
You have a Gemini Gem for your business. Gemini's share of assistant usage grew sharply through 2025 and 2026, and building only for ChatGPT now leaves a meaningful audience unserved. 2 points.
Your Instagram DMs get an intelligent response when you are unavailable. Through Meta AI or an equivalent, answering in your voice rather than a generic autoresponder. 2 points.
Every assistant you run routes people to a booking page or contact form. This is where most implementations fail. An assistant that helps beautifully and never directs anyone anywhere is a free help desk. 2 points.
Every enquiry, from any channel, lands somewhere trackable with a follow up date. Not scattered across an inbox, a DM folder and your memory. 2 points.

Layer 5 score: ___ / 10

Your total: ___ / 50

0 to 14: Invisible

You are not in the conversation. Assistants have almost nothing to work with, and any marketing spend is running into a hole. This is the most common band, and also the fastest to move out of, because the fixes are mechanical rather than creative.
Do first: Layer 1 in full, this week. Indexing before anything else.

15 to 24: Findable if someone already knows your name

Your site exists and can be reached, but you are a claim rather than a corroborated entity, and you are unlikely to be recommended to anyone who is not already searching for you specifically.
Do first: Layer 2, then Layer 4. Entity clarity plus corroboration.

25 to 34: Legible but not preferred

You would survive a comparison if an assistant found you. It usually will not, because competitors with stronger corroboration get selected first.
Do first: Layer 4 aggressively. Reviews and independent mentions are your bottleneck.

35 to 44: Competitive

You are genuinely in the running for specific, local, well phrased queries. The remaining gains are in Layer 5, which is where visibility converts into revenue.
Do first: publish and connect your assistants, and make sure every route ends at a booking.

45 to 50: Ahead of your market

Very few small service businesses are here in 2026. Your work now is maintenance and content depth: keep answering new questions, keep earning mentions, and review monthly what your assistants are telling people in your name.

The three most common failures we see

After running this audit repeatedly, three patterns dominate.

Zero Bing presence. Businesses submit to Google, stop, and never realise the most used assistant in the world is reading a different index.
Beautiful copy, no facts. Pages full of warm positioning that never state the city, the price, the timeline, or what is actually included. There is nothing for a machine to quote and nothing for a buyer to decide on.
Assistants with no routing. A business publishes a custom GPT, feels modern, and never connects it to a booking. Conversations happen, nobody is captured, and the effort produces no revenue. This is the failure that annoys us most, because the work was already done.

What this checklist deliberately does not promise

No provider, including us, can guarantee that your business will appear inside an AI assistant's recommendations. Those outputs are not controlled by anyone outside the platforms, and they change. Anyone selling guaranteed placement is overselling.

What you can control is everything on this list: whether you are indexed, whether you are legible as an entity, whether your content is quotable, whether independent sources corroborate you, and whether the attention you receive gets captured. Businesses that score well on all five layers show up far more often than those that do not. That is a strong correlation, not a guarantee, and the difference matters.

What to do with your score

If you scored below 25, start with indexing and entity clarity this week. Both are mechanical, both are free, and neither requires a decision about strategy.

If you scored above 25 and the gap is Layer 5, that is the layer we build. Clause & Code installs a complete AI client system for coaches, consultants and professional practices: an AI readable landing page, your own custom GPT, your Gemini Gem, Meta AI answering your Instagram DMs, and the routing that connects all of it to your booking. Fixed price, installed in days.

See what is included or book an intro call and we will run this audit with you properly.

AI Discoverability Checklist and Self Audit: Score Your Current Online Presence