Here's a question worth sitting with: when someone asks ChatGPT "what's the best [thing your company sells]," do you actually show up? Not on page two. Not buried in a list. Do you show up at all? Most brands genuinely don't know. They're obsessing over Google rankings while an entirely separate search ecosystem — one growing at 527% year-over-year — operates with zero measurement, zero optimization, and zero awareness on their end. That's the gap we're talking about today.

This isn't a post about AI SEO theory. It's about the practical, honest answer to "how do I know if AI search is working for me?" Because right now, the honest answer for most companies is: you don't. And that's a problem you can actually fix.

Why Your Google Rankings Tell You Nothing About AI Visibility

Traditional SEO metrics — keyword rankings, GSC impressions, organic clicks — measure one thing: how you show up in Google's ten-blue-links results. That's still important, but it's increasingly incomplete.

AI search platforms like ChatGPT, Perplexity, Gemini, and Google AI Mode don't operate on the same logic as keyword rankings. They're not showing you position 1 through 10. They're synthesizing an answer and either mentioning your brand or they're not. Your competitor might be getting recommended in thousands of AI conversations per day while ranking below you for every keyword you track. You'd never see it in your current dashboards.

Google Search Console's AI Impressions report helps somewhat — it shows you impressions from AI Overviews and AI Mode within Google's ecosystem. But it tells you nothing about what's happening in ChatGPT, Perplexity, Claude, or Gemini's standalone app. Those platforms don't send you impression data. You have to go find it yourself.

The Citation Gap Is Real

A study tracking 19 GA4 properties found that AI search drove a 527% year-over-year increase in referral traffic. But here's the uncomfortable flip side: the brands capturing that traffic represent a tiny fraction of businesses that should be visible. There's a massive citation gap — well-established brands, including some with strong Google rankings, that simply aren't being cited because they haven't structured their content the way AI models pull from sources.

The brands winning in AI search right now aren't necessarily the biggest ones. They're the ones that got there first — that got cited in training data, that got picked up by early AI indexes, and that showed up repeatedly in the kinds of sources AI platforms trust (Reddit threads, third-party reviews, industry directories, authoritative editorial coverage).

How to Actually Audit Your AI Search Visibility

Before you spend money on tracking tools, do this manually. It takes about 20-30 minutes and gives you a real baseline. Open ChatGPT, Perplexity, Google Gemini, and optionally Claude. Run three categories of prompts for each platform.

1. Branded Prompts

Ask directly: "What do you know about [Brand Name]?" and "Is [Brand Name] a good option for [use case]?" This tells you if AI platforms have knowledge of your brand at all, and whether that knowledge is accurate and positive. If ChatGPT says your company "appears to offer" something that you definitively sell, that's a red flag — it means the AI's knowledge of you is thin and potentially hallucinated.

2. Category Prompts

Search your category without your brand name: "What's the best [product/service] for [specific use case]?" and "Which companies do [thing you do]?" This is the most important test. Are you in the conversation when someone isn't specifically looking for you? Most brands aren't. If your competitor is in the top three mentions every time and you're not appearing at all, that's your competitive gap in plain sight.

3. Comparison Prompts

Try: "[Your Brand] vs [Competitor] — which is better?" and "Compare [your brand] and [competitor name]." How AI answers comparison queries tells you a lot about perceived authority. If the AI frames you as the underdog, or doesn't have enough information to compare fairly, that's actionable.

Do this across at least three platforms. ChatGPT and Perplexity will give you the most divergent results — they use different underlying models and different retrieval strategies, so your visibility in one doesn't predict the other.

💡 Quick Take

Run the same prompt five times in a row on Perplexity. It often returns slightly different citations each time. That variability tells you whether your brand is in the "confident mention" tier (shows up most times) or the "occasional mention" tier (shows up sometimes). Consistent mentions across multiple runs = you're in the model's core knowledge. Sporadic mentions = you're on the fringe of what it's pulling.

The Four Metrics That Actually Matter

Once you start auditing, you need a consistent way to score what you find. Here's a simple framework that mirrors what the dedicated tracking tools use:

Mention Rate

Out of 10 relevant prompts you test, how many times does your brand appear in the AI's answer at all? If you're mentioned in 3 out of 10, your mention rate is 30%. Track this per platform — your mention rate in ChatGPT and your mention rate in Perplexity are often very different numbers. A mention rate below 20% for your core category queries is a problem worth prioritizing.

Citation Rate

Perplexity and Google AI Mode show their sources. ChatGPT with web browsing enabled often does too. Citation rate is how often your actual website is linked as a source — not just a mention of your brand name, but a link back to your content. Citation rate is harder to achieve but more valuable. It means traffic. A brand can be mentioned without being cited, but being cited means someone can actually click through to you.

Share of Voice

When your category gets mentioned, what percentage of AI responses feature you vs. your competitors? If you're running 10 category-level prompts and ChatGPT names your competitors in 8 of them but only names you in 2, your share of voice is 20% vs. their 80%. That's a market positioning problem that shows up in AI before it shows up in revenue.

Answer Sentiment

This one's easy to miss. How does the AI characterize your brand when it does mention you? Is it positive, neutral, or hedged? Phrases like "some users report issues with" or "while popular, it's worth noting that" are signals that negative content in your brand's online footprint is influencing AI outputs. AI models are essentially aggregating what the web says about you — if reviews, Reddit threads, or press coverage skew negative, the AI answers will reflect that.

Tools Worth Using Once You've Done the Manual Audit

Manual auditing is great for a baseline but doesn't scale. If you're managing a brand that needs ongoing tracking, here's what's actually useful right now:

Profound (profound.io)

Built specifically for AI search visibility. It runs structured prompts across ChatGPT, Perplexity, and Gemini on a schedule and tracks your mention rate and share of voice over time. The reporting is clean and it's genuinely built around the metrics that matter, not retrofitted from SEO ranking tools. Starts at around $500/month, which is steep for smaller brands but reasonable for agencies or larger companies.

Rankscale

More affordable option, focused specifically on ChatGPT visibility tracking. Lets you define your target prompts, runs them repeatedly, and tracks how your brand shows up vs. competitors. Good if you're primarily concerned with ChatGPT and want something accessible without enterprise pricing.

Semrush's AI Toolkit (beta)

If you're already on Semrush, they've rolled out AI visibility features in the beta toolkit. It's not as purpose-built as Profound but it's convenient if you're consolidating tools. Worth turning on if you have access.

Manual Spreadsheet Tracking

Genuinely not a joke. For many smaller brands and agencies, building a 20-row Google Sheet with your top prompts, running them weekly in ChatGPT and Perplexity, and logging the results manually is more sustainable than paying $500/month for a tool you don't have the bandwidth to actually use. The discipline of a weekly manual check will catch most of what matters.

What Actually Improves Your AI Visibility

This deserves its own post (and we've covered parts of it in our Answer Engine Optimization guide and GEO strategy piece), but here's the short version once you've identified a visibility gap:

Bottom Line

If you haven't done an AI search visibility audit yet, go do it today. Open ChatGPT, run ten category prompts for your industry, and see how many times your brand shows up. The answer will either reassure you or light a fire under you — either way, you'll know something you didn't know before.

The brands that are ahead in AI search right now aren't ahead because they figured out some secret. They're ahead because they looked earlier. They measured something when it was still possible to move the needle without a huge budget. That window is still open, but it won't be forever — AI search visibility is already becoming competitive in high-value verticals, and playing catch-up a year from now will be significantly more expensive than acting today.

Measure first. Then fix what's broken. That's still the right order of operations.

🧪 Check Your Site's SEO Health

Run a free 60-factor audit on your site. No login, no credit card. Just results.

Run Free SEO Audit →