The DIY AI Visibility Check: Test How AI Sees Your Business
A step-by-step method to test whether ChatGPT, Perplexity, Gemini, and Copilot mention your business, score your Share of Answer, and fix the gaps. Published July 30, 2026.
A DIY AI visibility check means asking ChatGPT, Perplexity, Gemini, and Copilot the questions your customers ask, then recording whether your business is mentioned, how it is described, and which sources get cited. Run 10 to 15 buyer-style prompts monthly, score the percentage of answers that include your brand (your Share of Answer), and log the sources cited for competitors. The result is a baseline of brand presence in AI answers, not a traffic forecast, plus a prioritized fix list.
Your customers are already asking AI tools who to hire, where to go, and what to buy. Whether your business shows up in those answers is now measurable, and you can measure it yourself in about an hour.
One framing matters before you start. AI visibility is a brand presence metric, not a traffic strategy: AI platforms send far less referral traffic than Google organic, and the value of appearing in answers is credibility and consideration, not clicks.
This guide gives you the full method: which prompts to run, how to score the results, and how to turn a low score into a fix list.
What Are You Actually Testing For?
You are testing three things: whether your business appears in AI answers for questions in your category, how accurately it is described when it appears, and which sources the AI leans on.
The core metric is Share of Answer: the percentage of relevant AI-generated answers that mention your brand. If you run 12 buyer-style prompts and appear in 3 answers, your Share of Answer is 25 percent for that prompt set.
Accuracy matters as much as presence. An AI that names your business but gets your services, location, or hours wrong is working against you in the exact moment a potential customer is evaluating options.
Which AI Platforms Should You Test?
Test the four your customers most plausibly use: ChatGPT, Perplexity, Gemini, and Microsoft Copilot. Add Claude if your customers skew professional or technical.
Each platform behaves differently. Perplexity cites sources aggressively and refreshes from the live web, Gemini draws heavily on Google's index and Business Profile data, and ChatGPT blends trained knowledge with web search depending on the question.
That variation is the point of testing them separately. It is common for a business to show up consistently on one platform and be invisible on another, and the gap tells you where the fixable problem lives.
What Prompts Should You Run?
Write 10 to 15 prompts that mirror real buying questions, not vanity questions. "Best CPA for small business in Castle Rock" is a buyer prompt; "tell me about [your business name]" is a vanity prompt that only tests whether AI knows you exist.
Build the set from three groups. Category prompts ("best [service] near [city]", "who should I hire for [problem]"), comparison prompts ("[your category]: what should I look for", "compare top [service] providers in [city]"), and problem prompts phrased the way customers talk ("my [thing] is broken, who fixes that in [city]").
Keep the exact wording in a spreadsheet and reuse it every time. Changing the prompts between rounds destroys your ability to compare results month over month.
How Do You Score the Results?
Create a simple grid: prompts as rows, platforms as columns. For each cell record three things: mentioned yes or no, description accurate yes or no, and any sources the platform cited.
Your headline number is Share of Answer: brand mentions divided by total answers across the grid. A 12-prompt, 4-platform run gives you 48 answers, so 10 mentions is roughly 21 percent.
Log competitor mentions in the same pass. Knowing that a competitor appears in 60 percent of answers while you appear in 15 percent turns a vague worry into a specific gap, and their cited sources become your roadmap.
What Do the Cited Sources Tell You?
The citations are the most actionable part of the whole exercise. When AI answers cite Yelp, a trade directory, a local news story, or a competitor's FAQ page, it is showing you exactly which sources it trusts for your category.
Pattern-match across your grid. If Perplexity keeps citing the same two directories you are missing from, those listings are your fastest move. If competitors get cited for their own content, the gap is on your website: pages that answer buyer questions directly enough to be quoted by AI engines.
Local businesses should pay special attention to map and profile data. Gemini and Copilot lean on Google and Bing profile data respectively, so listing gaps show up fast in their answers. Our local business AI search guide covers that layer in depth.
How Often Should You Rerun the Check?
Monthly is the right cadence for most small businesses. AI answers shift as models update and as the underlying sources change, and monthly reruns are frequent enough to catch movement without turning into busywork.
Rerun the identical prompt set each time, on the same platforms, and date every round. Three or four rounds in, you have a real trendline: Share of Answer moving up or down, accuracy issues resolved or recurring, new competitors entering answers.
Expect noise. AI answers are not deterministic, and the same prompt can produce different mentions on different days. That is why the score is measured across a full prompt set rather than any single answer.
What Should You Fix First If Your Score Is Low?
Fix accuracy problems first, because a wrong description does active damage. Wrong data in AI answers usually traces to inconsistent listings, so a NAP and citation cleanup is typically the first move.
Second, close the source gaps your citation log revealed: claim the directories AI already trusts in your category, and strengthen the profiles it quotes from. Third, build content that answers your prompt list directly on your own site, structured the way AI systems extract and quote content.
Then rerun the check and let the trendline judge the work. Presence in AI answers builds credibility and consideration with buyers who never click anything, which is exactly why it is worth measuring even though it will never show up cleanly in analytics.
Frequently Asked Questions
AI Visibility Checks, Answered
How do I check if my business shows up in ChatGPT?
Ask ChatGPT the questions your customers would ask, such as "best [your service] in [your city]" or "who should I hire for [the problem you solve]," and note whether your business appears and how it is described. Run 10 to 15 of these buyer-style prompts rather than judging from one answer, because AI responses vary between sessions. Avoid asking about your business by name, since that only tests whether the model recognizes you, not whether it recommends you.
What is a good Share of Answer score?
There is no universal benchmark, because the score depends on your market, category competitiveness, and prompt set. The practical approach is to treat your first full check as the baseline and measure movement against it, while logging competitor mention rates in the same grid for context. A business appearing in under 10 percent of relevant answers while a direct competitor appears in half of them has a clear, specific gap to close.
Do I need paid tools to track AI visibility?
No. A spreadsheet, a fixed prompt list, and an hour a month cover the core method for a single-location or single-market business. Paid AI visibility platforms add automated tracking across many prompts, markets, and platforms, which starts to matter for multi-location brands or agencies managing several accounts. Start manual, and let the size of the job tell you if automation is worth it.
Why does my business show up in Perplexity but not ChatGPT?
Each platform builds answers from different sources: Perplexity searches the live web and cites aggressively, while ChatGPT blends trained knowledge with optional web search, and Gemini leans on Google's index and Business Profile data. Appearing on one platform but not another usually means the sources that platform trusts are covered while the other platform's sources have gaps. The citations in the answers where you do appear are the best clue to what the missing platforms want.
Does showing up in AI answers bring more website traffic?
Not meaningfully, and treating AI visibility as a traffic play sets the wrong expectation. AI platforms send a small fraction of the referral traffic that Google organic search delivers, and many buyers act on an AI recommendation without clicking anything. The value of AI visibility is brand presence, credibility, and consideration at the moment a buyer asks for recommendations, which is why Share of Answer, not traffic, is the metric to track.
How is an AI visibility check different from checking my Google rankings?
Rank tracking measures your position in a list of links for a keyword, while an AI visibility check measures whether you are named inside a synthesized answer to a buyer's question. AI answers pull from many sources at once, describe businesses in prose, and often recommend a short list rather than ten blue links. The two measures complement each other, and strong local rankings frequently feed AI answers, but neither substitutes for the other.
What if AI describes my business incorrectly?
Trace the error to its source rather than trying to correct the AI directly. Wrong hours, services, or locations in AI answers usually come from inconsistent or outdated listings, old directory entries, or stale website copy, and fixing those sources fixes the answers over the following weeks. Prioritize your Google Business Profile, Bing Places, Apple Business Connect, and the specific sources cited in the inaccurate answers.
About the author. Jaron Mossman is the founder of 360ROI, a boutique digital marketing consultancy based in Castle Rock, Colorado. He spent two years managing multimillion-dollar advertising accounts at Google's Manhattan office for Fortune 500 travel and hospitality brands before founding 360ROI in 2013. He has delivered AEO and GEO optimization as a live client service since 2024.