Test It Yourself: Ask ChatGPT Who Does SEO in Your Area.

By Khamare Clarke · August 2025 · 8 min read

You can do this right now, before you finish reading this post. Open a new tab. Go to ChatGPT. Type:

Then do the same on Gemini and Perplexity. You will get an answer. It will mention specific businesses. Read it carefully, because what you are looking at is not a chatbot's opinion. It is a commercial ranking. It reflects who has AI visibility in your market, and AI visibility is increasingly where high-intent buyers start their search.

If your business appears, that is a signal worth understanding and strengthening. If your business does not appear and your competitors do, you have a concrete gap to close. Either way, the three minutes this test takes will tell you more about your current market position than most quarterly reports.

Step-by-step: how to run the test properly

Step 1: Set up three tabs

Open the following in separate tabs:

• ChatGPT (chatgpt.com) — the most widely used generative AI platform. Use GPT-4o if you have access; the base model otherwise.
• Gemini (gemini.google.com) — Google's generative AI. This is particularly important because Google's AI Overviews in search results draw on similar data. Understanding your Gemini visibility gives you a proxy for AI Overview presence.
• Perplexity (perplexity.ai) — a retrieval-augmented search engine that cites sources in real time. Perplexity shows you both the answer and where it came from, which makes it uniquely useful for understanding what content is being cited.

Step 2: Run the right queries

Do not just ask one question. Run several variations that reflect how a real buyer thinks:

• “Who does roofing in Stoke-on-Trent?”
• “Best roofing companies near me” (with location enabled)
• “Recommend a reliable roofer in Staffordshire”
• “Who should I call for an emergency roof repair in the Midlands?”

The variation matters because different phrasing activates different retrieval pathways. A business that appears on the first query but not the fourth has partial AI visibility. Full AI visibility means appearing consistently across query variants, across platforms.

Step 3: Note who appears and where their citations come from

On Perplexity, you can see the source links. Click them. The sources will typically be: directory listings (Checkatrade, Yell, TrustATrader), local news articles, review platforms, the business's own website, or specialist trade publications. This tells you exactly where AI citation authority comes from in your market.

On ChatGPT and Gemini, you will not always see explicit citations, but the underlying mechanism is similar: the models surface businesses that have a strong, consistent presence in the text data they were trained on, combined with current web retrieval where RAG is active.

Record which businesses appear across all three platforms and all query variants. That list is your competitive AI visibility benchmark.

What does appearing in AI answers actually mean commercially?

In traditional search, appearing in position 1 on Google for a high-intent keyword is valuable because users click on it. The click-through rate at position 1 is documented at around 27% on average (First Page Sage, 2024), meaning roughly one in four people who see the result click through.

In generative AI, the dynamics are different. AI answers often do not generate a direct click. The user reads the answer, sees the business name recommended, and then separately searches for that business or calls directly. The commercial effect is a branded search attribution gap: you are driving conversion activity that appears in your direct or branded traffic, not in AI referral traffic. Most analytics setups do not capture this correctly.

What appearing in AI answers does mean concretely:

• Your business is named at the moment a buyer is actively in the consideration phase. That is the highest-value moment in the customer journey.
• Being cited by an AI system carries implicit endorsement. Most users do not question why an AI recommended a specific business. They treat it as a credible shortlist.
• Consistent AI citation builds brand recall. A buyer who sees your business named across ChatGPT and Perplexity and then searches Google and finds you in the map pack has had three touchpoints before making contact. Conversion rates from that kind of multi-touchpoint exposure are significantly higher than from a single search result.

What does not appearing mean?

It means your competitors who do appear are receiving the implicit endorsement you are not. In a local service market with three or four serious competitors, the one that appears in AI recommendations consistently has a structural advantage in the discovery phase of the customer journey.

It also means your entity data is insufficient. AI systems build their understanding of local businesses from entity consistency: whether your business name, address, phone number, and service description appear consistently across the web. Inconsistent NAP data, thin or absent third-party mentions, missing structured data, and low review volume all contribute to poor AI entity recognition.

The businesses that appear in AI recommendations are almost always those with: strong Google Business Profile signals, high review volume and recency, consistent directory listings, authoritative mentions on third-party sites, and well-structured website content that directly answers common service queries.

How is AI visibility actually built?

This is where the technical discipline of Generative Engine Optimisation (GEO) sits. GEO is not a single tactic. It is a combination of:

Entity consistency across the web

Every directory, review platform, and third-party mention of your business should have identical name, address, phone number, and website URL. Inconsistencies create entity disambiguation problems for AI systems trying to identify which “Clarke Roofing” is in Stoke and which is in Derby. Entity confusion means citation frequency drops.

Structured data implementation

Schema markup (JSON-LD) tells search engines and AI crawlers exactly what your business is, what services it provides, where it operates, and what its authority signals are. LocalBusiness schema, Service schema, FAQ schema, and Review schema all contribute to the structured entity profile that AI systems pull from.

This is not optional infrastructure for AI visibility. It is foundational. A site without well-implemented structured data is giving AI systems less to work with than a site that has it. In a competitive local market, that gap translates directly to citation frequency.

Citation in authoritative content

A 2024 study by Seer Interactive found that brands mentioned in authoritative third-party content were more likely to appear in AI-generated answers than brands present only on their own domains. This means your presence in local news, trade publications, case study features, and high-authority directories is an active AI visibility signal.

Getting mentioned in a Staffordshire business directory, a local construction industry publication, or a case study on a supplier's website all contribute to the web of citations that AI systems use to assess entity authority. This is a form of off-site SEO that has always mattered for Google rankings and now matters for AI visibility too.

Content that directly answers generative queries

Generative AI answers are built by assembling relevant content from across the web. If your website has a page that directly answers “what is the average cost of roof repair in Stoke-on-Trent?”, that page is a candidate for citation when someone asks ChatGPT the same question. If your site does not have that content, you are not a candidate.

This is why content strategy for GEO is different from traditional keyword-driven content strategy. You are not writing for a keyword. You are writing for a question format that a generative system will encounter from real users, and you are structuring your answer so that an AI system can cleanly extract and cite it.

What does the test tell you to do?

If you ran the test at the top of this post, you now have a real data point on your AI visibility position. Here is how to interpret what you found:

• You appeared consistently across all three platforms: Your entity data is in reasonable shape. The work is to understand why you appeared, systematise it, and expand it to more query variants.
• You appeared on one platform but not the others: You have partial AI visibility. The gap is likely entity consistency and structured data. You are in the system but not with enough signal strength to surface reliably.
• You did not appear on any platform: Your entity data is thin or inconsistent, your structured data is absent or minimal, and your third-party citation profile is insufficient. This is the starting point for a GEO engagement.

The test takes two minutes. Costs nothing. If you want to know your current AI Visibility Score in more detail, including a breakdown of where your entity data is inconsistent and which citations you are missing in your specific market, that is a conversation I run in the 30-minute call below.