How to Audit Your Competitors' Visibility in ChatGPT and Perplexity (Tactical Guide)

# How to Audit Your Competitors' Visibility in ChatGPT and Perplexity (Tactical Guide)

While you're not measuring, your competitors are. Last month one of your rivals discovered that ChatGPT recommends them first in 8 of the 12 questions their customers ask before buying. You don't know whether you appear in those responses at all.

This guide is the operational method: how to manually audit your competitors' visibility in ChatGPT and Perplexity right now, without any tools, in under two hours — and when it makes sense to move to automation.

Why Auditing Competitors in AI Is Different from Classic SEO Competitive Analysis

Classic SEO competitive analysis is clear: you check Google positions with SEMrush, calculate keyword overlap, compare backlinks. There's a stable SERP and mature tooling.

In GEO things change. There's no SERP — there are synthesized responses. The same question worded differently produces different answers. And models change behavior week to week. Competitive analysis here isn't about comparing rankings; it's about comparing frequency, position, and context within conversational responses.

Three practical consequences:

• It's not enough to look at your own brand. You need to measure the entire competitive universe within each prompt.
• The key metric is not individual position, it's relative share of voice: what percentage of the relevant responses in my category include me versus the percentage that include competitor X?
• Sentiment matters as much as presence. Appearing frequently with negative framing is worse than not appearing at all.

Three Things You Need to Know Before You Start

1. Your competitor set isn't your commercial hit list. Your competitors in GEO are the brands that appear alongside yours in AI responses. Sometimes they're not the ones your sales team considers competition. A new entrant in your category that you've never tracked could be eating your share of voice. You only discover that by measuring.

2. Same prompt, different session, different response. ChatGPT has memory across sessions for the same user. If you're logged in and have previously asked questions related to your brand, the responses to your audit prompts are biased. Always audit from clean sessions — incognito mode or an account with no relevant history.

3. One model doesn't give you the full picture. ChatGPT, Perplexity, Gemini, and Claude respond differently. Auditing only ChatGPT leaves you blind to 50–60% of real buyer behavior. A minimum viable audit covers at least three models — know the differences between ChatGPT, Gemini, and Perplexity before you start.

10 Prompts to Manually Audit Your Competitors

Replace `[your category]` with your vertical (CRM, SEO agencies, accounting software, boutique hotels in Austin — whatever applies) and `[your brand]` with your company name. Run each one in an incognito session, in both ChatGPT and Perplexity.

Prompt 1. Pure discovery.

"What are the best [your category] in 2026?"

What you're looking for: a neutral list. Note which brands appear, in what order, and whether yours is included.

Prompt 2. Three-way comparison.

"Compare [your brand] with [competitor 1] and [competitor 2]. Which is better for [typical use case]?"

What you're looking for: how the AI describes you against those two rivals. Sentiment, strengths, weaknesses.

Prompt 3. Use-case recommendation.

"I'm [typical buyer profile] and I need [specific problem solved]. Which [your category] do you recommend?"

What you're looking for: your position on the list for your exact ICP. This is where you discover whether AI recommends you to your ideal customer or not.

Prompt 4. Pricing.

"Which [your category] is the most affordable, and which is the premium option?"

What you're looking for: does AI correctly classify you within your price range, or does it place you somewhere you don't want to be?

Prompt 5. Alternatives to the leader.

"Give me 5 alternatives to [category leader]"

What you're looking for: if you appear in the alternatives list for the market leader, you're considered a real option. If not, you're invisible to everyone evaluating the leader.

Prompt 6. Reverse angle.

"Give me 5 alternatives to [your brand]"

What you're looking for: the AI tells you who your real competitors are in its model of the world. Sometimes brands you've never tracked appear here — a shortcut for competitor discovery that wasn't on your map.

Prompt 7. Geography.

"Which [your category] is best for companies in the US / UK / [your market]?"

What you're looking for: if you sell locally, this prompt reveals whether AI associates you with the right market or ignores you in favor of global brands.

Prompt 8. Feature-specific.

"Which [your category] has the best [your key differentiating feature]?"

What you're looking for: does AI associate you with the feature that is your core differentiator? If not, your positioning hasn't reached the model.

Prompt 9. Reviews.

"What do users say about [your brand]? And about [competitor 1]?"

What you're looking for: sentiment. Compare positive versus negative framing in both cases. Identify which source the model is drawing those opinions from.

Prompt 10. Active citation.

"Recommend a [your category] and give me a link to their website."

What you're looking for: citation rate. If AI recommends your competitor with a link and you without one, you have a schema markup and authority problem.

A thorough manual audit is these 10 prompts × 3 models = 30 runs. It takes 90 minutes the first time, 45 once you have a template.

How to Interpret the Results: A 4-Quadrant Framework

When you finish the 30 prompts you'll have a messy matrix. To extract signal, use this framework: classify each competitor on two dimensions.

X-axis: appearance frequency (in how many of the 30 prompts did they appear)

Y-axis: quality of appearance (average sentiment + average position)

Four quadrants:

• Established leaders (high frequency + high quality): the benchmark competitor. If you're not here, this is your 12-month target.
• Volume without substance (high frequency + low quality): AI cites them often but criticizes them. Vulnerable. If you're nearby, this is where you can gain ground quickly.
• Strong but niche (low frequency + high quality): they appear infrequently but well. Usually a very strong vertical niche. Study what they do.
• Invisible (low frequency + low quality): not real GEO competition, even if they compete with you in your offline market.

Where do you sit? That's the first question. The second: what move shifts my quadrant?

From Manual to Automated: When to Make the Jump

Manual auditing works for an initial snapshot and for validating hypotheses. It doesn't work for continuous tracking. Three clear signals that it's time to automate:

1. You want trend, not snapshot. A manual audit tells you where you stand today. It doesn't tell you whether you're moving up or down. To detect trends you need to run the same prompt set every week and compare deltas. Impossible by hand at 30+ prompts.

2. Your category moves fast. If your vertical sees 2–3 new competitors per year or models change behavior monthly, a manual snapshot ages in weeks. You need automatic updates.

3. You want alerts, not reports. A competitor sits in position 1 one week and drops to position 5 the next. That's a signal something changed: schema, content, reviews. Detecting that without automated alerts is flying blind.

When you make the jump, the right setup measures the 10 prompts expanded to 30–50, across 4–5 models, weekly, with tracking of every key metric per competitor. See how Mentio automates this.

5 Mistakes That Invalidate a GEO Competitive Analysis

1. Auditing while logged in. If you run the 30 prompts from your personal ChatGPT account with history about your brand, results are biased. Incognito mode, always.

2. One prompt per idea. "Best CRM?" and "What's the best CRM for my business?" return different responses. Audit variants, not just the canonical phrasing.

3. Ignoring the second paragraph. AIs typically put obvious recommendations in the first paragraph and nuances and alternatives in the second and third. That's where brands that aren't top-of-mind appear. Read the whole response.

4. Mixing languages. ChatGPT in English and in Spanish cites different brands. If your market is English-speaking but you audit in another language, the results are irrelevant to your case.

5. Auditing without a baseline. Without a first measurement and a fixed prompt set, you can't know whether you improved. Define your baseline before making any changes to your content.

Case Study: Spanish SEO Agency Discovers It's Invisible to Its ICP

A B2B SEO agency based in Madrid ran this exercise in February 2026. They audited 10 prompts × ChatGPT, Perplexity, and Gemini.

Result: they appeared in 4 of the 30 prompts executed, all in positions 5–7. Their three main competitors appeared in 18–22 prompts, several in positions 1–2. AI described those competitors as "leaders in technical SEO for mid-sized companies in Spain"; the audited agency wasn't even associated with that specialization, even though it was their core offering.

Root cause after investigation: two competitors had 30+ published posts on technical SEO for enterprise clients; the audited agency had 8 scattered posts. AI learns from the mass of thematic content, not from isolated quality.

Action taken: a cluster of 20 posts on enterprise technical SEO over 6 months + Organization schema + outreach to industry publications.

Result after 4 months: appears in 14 of 30 prompts, average position 3.4. Not a leader yet, but now a real option. Next quarter targets a consolidated top-3 position.

FAQ

Is it legal to audit competitors in ChatGPT? Completely. You're asking questions to a public AI, just like any other user. You're not accessing any private data belonging to a competitor — you're only observing how the AI publicly talks about them. It's exactly the same principle as looking at Google search results.

How many times should I repeat each prompt? At least twice in separate sessions. AIs introduce stochastic variability: the same prompt can produce slightly different responses. With two runs you identify whether there's a stable pattern or whether the response is noise.

Is running the audit only in ChatGPT enough? It works as a starting point but it's incomplete. Perplexity has higher commercial intent per session, Gemini is growing in B2B Google Workspace environments, and Claude dominates in advanced technical use cases. Covering only ChatGPT leaves you blind to at least one of your buyer segments.

How quickly can a competitor's AI visibility change? Between 4 and 16 weeks if they execute well (content + schema + authority). Dramatic position shifts from one month to the next usually indicate a competitor launched a PR campaign or published an important content cluster. Worth investigating.

What should I do if a competitor far outpaces me in share of voice? First, determine whether they're a real competitor or just have a longer web presence. Second, study which sources AI uses to talk about them. It's usually their own content cluster plus 2–3 third-party publications. Replicate the pattern in your own category.

Can I automate this with custom scripts, or do I need a dedicated tool? Technically yes, using the OpenAI and Perplexity APIs. Expect 40–60 hours of development plus ongoing maintenance. For 1 brand it makes sense if you have spare dev capacity; for 5+ brands a specialized tool almost always costs less than your own build.

Start Your Competitive Audit with Mentio

Mentio automates the 10 prompts expanded to 50, across ChatGPT, Gemini, Perplexity, Claude, and AI Overviews, with weekly tracking and alerts whenever a competitor moves their position.

See your competitive radar for free →

---

Want to know if AI mentions your brand?

Discover your visibility in ChatGPT, Claude, and Gemini in minutes.

Analyze your brand for free

Want to know if AI mentions your brand?

Discover your visibility in ChatGPT, Claude and Gemini in minutes.

Analyze your brand for free

Related articles