Why checking manually matters — and why one query isn't enough

Before we get into the steps, let's address a misconception that trips up a lot of business owners: running one query into ChatGPT and checking whether your name appears is not an AI visibility audit. It's a data point. A single query in a single AI tool gives you a narrow, unreliable snapshot of your actual visibility.

Here's why: AI systems don't produce identical outputs every time. They're probabilistic, meaning the same question can yield different answers on different occasions. Your business might appear in one response and not in the next. Additionally, different query formats trigger different recommendation behaviours — "who's the best plumber near me?" might get a different answer than "I need to find a reliable plumber in downtown Austin, any recommendations?"

To get a meaningful picture of your AI visibility, you need to test across multiple queries, across multiple platforms, and ideally repeat the process every 30–60 days to track changes. That's what this guide teaches you to do.

Step 1: Define your buyer-intent queries

The single most important step is often skipped: you need to think carefully about what your customers are actually asking AI systems before you start checking anything.

Your customers don't ask AI "what is the SEO optimised keyword for plumbing services in Austin TX." They ask the way they talk to a knowledgeable friend. Your queries need to mirror that natural language.

The core query formula

Start with this pattern and build variations around it:

"[best/top/recommended] + [your service] + [in/near] + [your city]"

For a dentist in Chicago, that produces queries like:

  • "What are the best dentists in Chicago?"
  • "Can you recommend a top dentist near downtown Chicago?"
  • "Who are the most recommended dentists in Lincoln Park?"

Add problem-based queries

Increasingly, people ask AI about their problem first and let the AI recommend a solution. Include queries that mirror how customers think about their problem, not just about your category:

  • "Who do I call when my furnace stops working in Minneapolis?"
  • "My kitchen drain is blocked — who should I call in Seattle?"
  • "I need a root canal and I'm terrified of dentists — who's good with anxious patients in Portland?"
  • "My roof is leaking and I need someone fast — who do I call in Phoenix?"

Add context and qualifier queries

Add variations with qualifiers that your specific customers might use:

  • "Best [service] for [specific type of customer] in [city]" — e.g., "best dentist for kids in Tampa"
  • "Most affordable [service] in [city]" — e.g., "most affordable electrician in Nashville"
  • "[Service] open on weekends in [city]"
  • "[Service] that accepts [insurance type] in [city]"

Your query target: 10 different queries

Aim for at least 10 distinct queries across these categories. Write them down before you start testing — you'll use the same list across all three AI platforms and again next time you check, so you want a consistent set you can compare over time.

Step 2: Check ChatGPT

ChatGPT is the most widely used AI assistant, making it the most important platform to check. Here's how to do it properly.

What to use

Use ChatGPT at chat.openai.com. The free version (GPT-4o mini) is sufficient for this check. If you have a paid subscription, use GPT-4o for potentially more current and detailed results.

Important: start a fresh conversation for each query

ChatGPT's responses are influenced by previous messages in the same conversation thread. If you ask 10 questions in a row, early answers can bias later ones. For accurate results, start a new chat for each query. It takes an extra 30 seconds and makes your results meaningfully more reliable.

How to phrase your queries to get business recommendations

ChatGPT sometimes hedges on local business recommendations — it may say it can't browse the internet or that it doesn't have current local listings. To get past this, be explicit about what you want:

  • Instead of "who is the best dentist in Austin?" try "Can you recommend specific dentists in Austin, Texas with strong reputations for their work?"
  • If it hedges, follow up with: "I understand you may not have current information — but based on what you know, are there any dentists in Austin you're aware of with strong reputations?"
  • Try both question formats to see if one yields more specific responses than the other

What to record

For each query, note: Did ChatGPT name specific businesses? If yes, which ones? Was your business named? Was it named first, in the middle, or last? Did ChatGPT say anything specifically positive or negative about any business it named?

Step 3: Check Perplexity

Perplexity (perplexity.ai) is an AI search tool that's designed specifically for finding current information. Unlike ChatGPT's base model, Perplexity actively searches the web for each query and cites its sources. This makes it a particularly important platform to check — and a different kind of check than ChatGPT.

Why Perplexity is different from ChatGPT

When Perplexity answers a local business query, it typically pulls results from review platforms, directories, and local publications that it finds via live search. Your AI visibility on Perplexity is closely tied to your presence on the platforms Perplexity is likely to surface — primarily Yelp, Google, and industry-specific directories.

How to check on Perplexity

Enter each of your 10 queries directly into Perplexity's search bar. After each answer, look at the source citations it shows. Note which platforms it drew from and which businesses it named. If your business appears, also note whether Perplexity cited a specific review platform or directory where it found information about you.

What to look for

On Perplexity, you're looking for: whether your business is named in the response, which sources Perplexity used, and whether your competitors have stronger representation in those sources. The source analysis is particularly useful — if Perplexity keeps citing Yelp and you have a weak Yelp presence, that's a clear action item.

Step 4: Check Google AI Overviews

Google AI Overviews (formerly Search Generative Experience) appear at the top of Google search results for many queries — a box of AI-generated information that synthesises multiple sources before the traditional search results appear.

How to trigger AI Overviews

AI Overviews don't appear for every query. They're most common for informational and local recommendation queries. To check them, open Google in a browser (not the Google app), type your query as a full question or natural language phrase, and look for the coloured "AI Overview" box at the top of results. If it doesn't appear for one query, try the next.

Note: AI Overviews are rolling out at different rates and may not appear for every location or every query type. If you don't see them for most queries, that's fine — check more carefully in 30–60 days as their coverage continues to expand.

What to record from Google AI Overviews

Note whether an AI Overview appeared, what it said, whether your business was named, what sources it cited (usually visible as small icons below the overview text), and whether competitors were named. Also note whether the AI Overview linked to review sites or your website directly.

Appeared in 0–3 of 10 queries (per platform) Poor — significant work needed
Appeared in 4–6 of 10 queries (per platform) Moderate — foundation exists, build on it
Appeared in 7–9 of 10 queries (per platform) Good — maintain and expand
Appeared in 10 of 10 queries (per platform) Excellent — focus on leading position
Named first in most recommendations Strong signal of AI authority
Named but not first, or inconsistently Presence established, room to grow

Step 5: Check your competitors

Checking your own visibility without checking your competitors' gives you half the picture. The other half — who is appearing instead of you, and what makes their digital presence different — is just as valuable.

Run the same 10 queries on all three platforms and track which other businesses appear. For the top two or three competitors you identify, ask yourself:

  • How many Google reviews do they have versus you?
  • Are they listed in directories you're not?
  • Do they have more detailed, substantive website content?
  • Have they been mentioned in any local news or editorial coverage?
  • Do their Yelp, Houzz, or industry-specific profiles look more complete than yours?

This isn't about feeling bad about where you stand — it's about identifying the specific gaps you need to close. If a competitor with 300 reviews is consistently beating you and you have 40, the path forward is clear. If a competitor is in directories you're not, that's a straightforward fix.

Sometimes you'll discover that the businesses dominating AI recommendations in your market aren't even your main competitors in the traditional sense — they might be larger, better-funded, or simply more digitally active. That's useful context too.

Step 6: Document your baseline

All of this information is only valuable if you write it down in a way you can compare to next month's check. Here's the tracking structure to set up — you can do this in a simple spreadsheet.

Your AI visibility tracking spreadsheet

Create a spreadsheet with the following columns:

  • Query: The exact question you asked
  • Platform: ChatGPT, Perplexity, or Google AI Overview
  • Date: When you ran the check
  • Your business appeared: Yes / No
  • Position: First mentioned, second, third, or not mentioned
  • Competitors named: List the businesses that did appear
  • Notes: Anything notable — was there a specific reason given, a citation to a particular platform, positive or negative language used?

Fill this in for each of your 10 queries across each platform — that's potentially 30 rows per check session (10 queries x 3 platforms). It takes 45–60 minutes total, and the information you gather is worth far more than the time invested.

Save this spreadsheet somewhere you'll find it again. Add a calendar reminder to repeat the process in 30–60 days. Over time, this document becomes a clear record of whether your AI visibility is improving, staying flat, or declining.

Why you should re-check every 30–60 days

AI systems change constantly. New training data gets incorporated, browsing capabilities are updated, weighting of different signals shifts. A check you did three months ago is a historical snapshot, not current reality.

Beyond the platforms themselves changing, your own digital presence changes — you get more reviews, you update your website, you get mentioned in a news article. Each of these changes can meaningfully affect your AI visibility, and you want to be able to see that progress in your tracking data.

The businesses that gain the most from AI visibility work are the ones tracking it consistently. They can see what's working, double down on it, and adjust when something isn't moving the needle.

The manual check you just learned is genuinely valuable — and it's also limited in important ways. You're testing a sample of queries, not the full range of ways customers might ask about your category. You're checking manually rather than at scale. And you're seeing the output but not the "why" — what specifically is causing your competitor to be recommended more often than you. A professional AI visibility audit adds competitive analysis, a full action plan, and ongoing monitoring. The DIY check is the right place to start; think of it as the check-up that tells you whether you need to see a doctor.

When to DIY versus when to use a professional check

The manual check we've described is a great starting point. It costs nothing, takes under an hour, and tells you the most important thing: whether you have an AI visibility problem worth addressing.

Here's an honest breakdown of what each approach gets you:

DIY manual check — best for

  • Getting a quick initial read on whether you appear in AI recommendations at all
  • Understanding which competitors are appearing in your place
  • Tracking month-over-month changes at a basic level
  • Deciding whether to invest more in this area

Professional AI visibility audit — best for

  • Businesses that have identified a gap and want to know exactly what to fix and in what order
  • Competitive markets where multiple businesses are fighting for AI recommendations
  • Getting a full technical analysis — robots.txt, schema markup, citation gaps, content issues — not just whether you appear
  • Receiving a prioritised action plan that takes into account your specific situation, not just generic advice
  • Ongoing monitoring across a much larger set of queries than you can run manually

The honest advice: do the DIY check first. It costs nothing and takes an hour. If what you find is concerning — you're consistently absent while competitors are named frequently — that's when a professional audit pays for itself, because it tells you exactly how to close the gap.

If you'd rather skip the manual work and get a professional check done right now, the free check available on this site is designed exactly for that: a clear, plain-English assessment of where you stand across the major AI platforms, with no jargon and no obligation to buy anything.