A quiet debate is playing out among the companies that sell "AI visibility" tracking — tools that tell businesses how often they get mentioned when people ask chatbots for recommendations. The question: should Perplexity, a smaller AI search engine, count as much as ChatGPT, Gemini, and Claude when calculating a brand's overall score?
The issue comes down to weighting. Visibility trackers typically run the same set of test queries across several AI platforms, then average the results into one composite score. If Perplexity handles a small fraction of real-world AI search traffic compared to ChatGPT, but gets equal billing in the average, a business's score can swing based on how it performs on a platform almost none of its customers actually use.
This isn't a product launch or a new feature. It's a methodology adjustment happening inside the analytics layer that's sprung up around AI chatbots over the past two years. Companies building these dashboards are essentially recalibrating their math as real usage data comes in, rather than relying on assumptions made when the tools first launched.
The underlying numbers explain the pressure. ChatGPT has several hundred million weekly users and is increasingly used for product research and shopping-style queries. Gemini rides on Google's default placement across Android and Search. Claude has carved out a foothold in business and technical use cases. Perplexity, while popular with a devoted niche and easy for vendors to query through its API, represents a much smaller slice of everyday consumer search behavior.
This pattern mirrors what happened with web analytics and social media metrics in earlier tech cycles. Early measurement tools tend to treat every platform as equally important because that's the easiest way to build a dashboard. Over time, as real usage data accumulates, vendors quietly adjust their formulas to reflect where attention actually concentrates. Vanity metrics get demoted; the loudest platform doesn't always stay the most measured one.
For small businesses, the practical risk is misallocated effort. If a visibility report shows a weak score driven mostly by poor performance on a platform with limited reach, a business owner might spend hours optimizing content for the wrong audience. Conversely, ignoring a platform entirely because it's small today ignores that AI search share can shift quickly as browsers, partnerships, and default integrations change.
The more useful move this week is checking your own site analytics rather than trusting a vendor's composite score outright. Most analytics platforms now show referral traffic broken down by source, including chatgpt.com, perplexity.ai, and gemini.google.com. Comparing actual visits from each platform against how a visibility tool weights them will tell you whether that tool's methodology matches your real customer base — or whether you're chasing a score that doesn't reflect where your buyers actually search.
It's also worth asking any AI visibility vendor directly how they weight platforms and whether that weighting updates as usage data changes. A tool that treats every chatbot as equally important, with no adjustment mechanism, is measuring something closer to "how well do you rank across every AI tool that exists" rather than "how visible are you to your actual customers."
Watch for three things over the next few months: whether Perplexity's usage grows meaningfully through its Comet browser or new hardware partnerships, whether OpenAI and Google expand shopping and local-business features inside ChatGPT and Gemini that make those platforms more commercially relevant, and whether visibility-tracking vendors publish transparent weighting methodologies rather than black-box composite scores.
The bottom line: AI visibility scores are only as useful as the assumptions baked into them, and those assumptions are currently in flux. Before acting on any vendor's dashboard, check it against your own referral traffic and ask how the underlying platforms are weighted.