← Blog

Does ChatGPT Know Your Business? Check in Five Minutes

· · 8 min read

Ask ChatGPT, Claude, Perplexity, and Gemini the same questions your buyers ask: category recommendations, branded queries, and comparison prompts. Run each in a private session so personalization does not skew the result. If your business does not appear, your buyers cannot find you through the channel half of U.S. adults now use first.

A business card illuminated by a searchlight beam in a dark void, Craig Pretzinger.

Open ChatGPT right now and ask it who the best provider is in your category. Type the question the way a real buyer would, not the way you would type it.

If your business does not appear in the answer, every buyer who asked that question today got steered toward someone else. That is the diagnostic. It takes five minutes. Most owners never run it.

TL;DR

Open four AI platforms: ChatGPT, Claude, Perplexity, and Gemini. Use a private session on each. Run three prompt types your buyers actually ask: a category recommendation, a branded question about your company, and a comparison question.

Record what each platform says about you and whether competitors appear. Run each prompt a few different ways because the answers change. This snapshot reveals whether your buyers can find you through the channel they are increasingly using first.

Why does this matter right now?

Half of U.S. adults now use AI chatbots, up from a third in 2024, according to Pew Research Center's June 2026 survey. During the same period, the share of Google searches ending without a click reached 68 percent. Two lines on a chart moving in opposite directions.

We covered the buyer behavior shift in detail in our referral pipeline analysis. The short version: your buyers stopped Googling and started asking. The businesses named in those answers get the call. The ones not named never knew they were in the running.

The platform landscape is also fragmenting. ChatGPT's share of B2B AI referral traffic fell from 89 percent in mid-2025 to 63 percent by early 2026, per Goodie's 2026 AI Search Traffic Report, while Claude grew from 1.4 percent to 18.5 percent. Gemini quadrupled and Perplexity more than doubled. Checking only ChatGPT means you miss more than a third of the AI referral landscape.

Which three prompt types reveal your AI visibility?

Your buyers ask different questions at different stages. A single prompt will not capture your full visibility picture. PowerChord's guide confirms that phrasing changes the answer, so run each prompt a few different ways across all four platforms.

What do category buyers ask?

Ask the AI to recommend a business in your category and market. Use the words your buyer would use, not industry language. If you run a commercial roofing company in Denver, type "who is the best commercial roofer in Denver." If you are an M&A advisor for dental practices, ask "who helps dentists sell their practice." Try at least three variations of the same question, changing one word each time to see if the answer shifts.

SparkToro's January 2026 research found less than a 1 percent chance of receiving identical brand recommendations across repeated queries. A single run is a single data point, never a score.

What does the AI say about your business by name?

Ask about your business directly. Type "What is [Your Company]?" or "Tell me about [Your Company] in [Your City]." This surfaces what the AI has assembled about you from whatever it has scraped.

The answers often reveal inconsistencies you did not know existed online: an old address, a service you stopped offering, a competitor's phone number. The AI assembles its answer from scattered sources. When your information appears in conflicting versions across those sources, the AI produces a composite that is wrong.

What comparison questions do buyers ask before deciding?

Ask the questions that happen right before a decision. "Who is better, Company A or Company B?" or "Is Company X worth the cost?" Your buyers are asking these right now. If a competitor appears in the comparison but you do not, the AI is steering the decision toward them before the buyer ever speaks to you. This is the hardest category to fix but the most important to know about because it tells you who the AI sees as your category default.

How do you read the results without fooling yourself?

The audit tells you what the AI said on that day, on that query, on that platform. Do not treat it as a pass or fail test. What matters is the pattern across platforms and prompt types. Here is what you are looking for, in order of severity.

FindingWhat it meansFirst action
You do not appear anywhereEntity signal problemAdd Organization schema. Standardize your listings.
You appear but are described wrongConsistency problemFind every online listing and make them match.
Competitors appear instead of youConsensus problemEarn mentions in authoritative external sources.
You appear on one platform but not othersPlatform retrieval gapIdentify missing sources and get listed there.

TrySight's research on how AI chatbots select brands confirms that every external source describing your business should reinforce the same core signals: your category, the problem you solve, and who you serve. When the model sees consistent answers across multiple sources, it gains confidence in the recommendation.

What should you do after the diagnostic?

Most business owners see a competitor's name and want to fix everything at once. Do not. The most common failure mode is skipping the mechanical basics because they feel too small. Here is the order to follow.

  1. Add Organization schema markup to your website. OpenAI's crawler documentation confirms OAI-SearchBot surfaces sites in ChatGPT search results. Structured data gives the crawler parseable signals about your name, category, location, and contact information in a format it can read without guessing.
  2. Standardize every online listing. Google Business Profile. Industry directories. Social profiles. Your website footer. Every instance of your business name, category, and location must match. A single variant in a single directory creates a fork in the model's understanding of what you are.
  3. Build third-party corroboration. This is the slowest lever but the highest ceiling. Omnia's practical playbook identifies authoritative off-site mentions as the signal that moves the needle most. Industry publications, trade association directories, and news mentions all contribute to the consensus layer that makes the AI confident enough to name you.
  4. Repeat the diagnostic monthly. The landscape shifts fast. A platform that ignores you today might be your best referral source in three months. The SBA's market research guide notes that competitive analysis is not a one-time event. The same holds for AI visibility.

We broke down the full mechanism behind entity recognition and retrieval in our post on how ChatGPT picks businesses. We also published a companion audit guide with detailed prompt templates for each platform. The diagnostic tells you the starting position. Those posts explain how the game is scored and how to run a full audit.

Sources cited in this analysis?

  1. SparkToro - AI Brand Recommendation Inconsistency Research 2026 - Under 1% chance of identical brand recommendations across repeated AI queries
  2. OpenAI - Overview of OpenAI Crawlers - OAI-SearchBot documentation and structured data guidance
  3. Pew Research Center - Americans and AI 2026 - Half of U.S. adults now use AI chatbots
  4. Goodie - 2026 AI Search Traffic Report - ChatGPT referral share fell from 89% to 63% in eight months
  5. Omnia - How to Improve Brand Visibility in ChatGPT 2026 - Practical playbook for systematic AI visibility work
  6. PowerChord - How to Check AI Search Visibility - Free AI Visibility Report methodology and prompt types
  7. TrySight - How AI Chatbots Select Brands 2026 - Ranking logic behind AI recommendations and brand selection

Frequently Asked Questions

Is one check enough to measure my AI visibility?

No, one check is not enough to measure your AI visibility. SparkToro found under a 1 percent chance of identical brand recommendations across repeated AI queries. Run the diagnostic monthly and across all four major platforms to see the real pattern rather than trusting a single snapshot that can mislead you.

Do I need paid tools for this diagnostic?

No, the three-prompt method costs nothing and takes about five minutes. Paid tools add repetition and scale for ongoing monitoring across hundreds of queries, but the first diagnostic requires only a browser, a private session, and the questions your buyers already ask.

What if the AI says my business does not exist?

The model cannot place your entity in its category. Add Organization schema to your website, standardize every online listing so your name and category match, and build third-party mentions across authoritative sources. This is a mechanical problem with a mechanical fix, not a permanent verdict.

Should I check only ChatGPT for AI visibility?

Check all four major platforms. ChatGPT's share of B2B AI referral traffic dropped from 89 percent to 63 percent in eight months according to the Goodie report. Claude, Gemini, and Perplexity now hold nearly 37 percent combined, and each platform uses different retrieval sources.

2026-07-18 - v4.0.0 - v4 conformant - built on The Standard