Why the average brand scores 31/100 and top brands 80+: reading an AI visibility report

The average brand we've measured scores 31 out of 100 on AI visibility. Top brands score 80+. The gap is not about budget or brand age. It's about what AI engines can find, parse, and cite when someone asks them a buying question.

We build Be Recommended at Inithouse. It scores how ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews recommend a brand, from 0 to 100, across 50+ real prompts. After running reports for dozens of brands, this is how we read them and what we'd fix first.

What makes up the score

Each report fires 50+ prompts at five AI engines. The prompts simulate how people actually ask AI for recommendations: "best X for Y," "alternatives to Z," "which tool do I use for..."

For each response, the system checks several things: is the brand mentioned at all? Is it recommended as one of the options? Is it positioned as a top pick? Is it cited with a working link? Each check contributes points. The per-engine scores roll up into a composite 0 to 100.

A brand scoring 31 typically means it gets mentioned in some responses on one or two engines, but it's absent from most prompts and most engines. A brand at 80+ gets mentioned across most prompts on most engines, often as a top recommendation with a source link.

How to read the per-engine breakdown

The composite number is useful for benchmarking, but the per-engine breakdown is where you actually learn something.

Here's a pattern we've seen repeatedly: a brand scores 60 on Perplexity (which favors well-cited, structured content) and 8 on Claude (which draws more from its training data). Same brand, same day, wildly different visibility.

Only about 11% of domains we've measured get cited by both ChatGPT and Perplexity. The engines don't share retrieval logic. They have different training data cutoffs, different citation habits, and different preferences for what counts as a trustworthy source. A high score on one engine tells you almost nothing about the others.

When you see a per-engine gap like 60 vs. 8, the question shifts from "how do I improve my AI visibility" to "what does Claude's training data lack that Perplexity's retrieval found?"

What separates 31 from 80

We've looked at dozens of reports now. The pattern in the 80+ scores is consistent:

Structured, factual content. Brands that score high tend to have comparison pages, detailed feature lists, and FAQ sections that directly answer the kinds of prompts AI engines use. Not marketing copy. Structured, parseable information that an AI can extract and cite.

Third-party mentions. A brand that only describes itself on its own website will struggle. Brands at 80+ show up in independent reviews, community discussions, technical comparisons on platforms like Reddit, Stack Overflow, or industry publications. The AI engines weigh third-party mentions heavily.

Consistent naming. Brands that use different names or abbreviations across different platforms confuse retrieval. If your product is called "DataSync Pro" on your website, "DSP" in community forums, and "Data Sync" in press releases, the engines treat these as potentially different things.

Recent, indexable citations. Perplexity in particular relies on fresh, crawlable sources. A brand with a blog post from 2024 mentioning its features will score differently than one with a technical comparison published last month. Recency matters, but only if the content is structured enough for retrieval.

What to check first

If you've never run an AI visibility audit, start with three questions:

1. Are you mentioned at all? Run a few buying-intent prompts through ChatGPT and Perplexity yourself. Ask "best [your category] tools for [common use case]." If your brand doesn't appear in any of them, you have a discovery problem, not an optimization problem.

2. Which engines know you? The per-engine gap is the most actionable signal. If Perplexity cites you but Claude doesn't, the issue is likely training data coverage, not content quality. If Claude mentions you but Perplexity doesn't, you might lack fresh, crawlable sources for retrieval.

3. Are you cited or just mentioned? There's a difference between "Brand X is an option" and "Brand X [link] is a tool that does Y." The cited version carries more weight in how AI engines present recommendations. Brands with structured, attributable claims get cited more often.

The 11% citation overlap problem

One stat from our data that keeps coming up: only about 11% of domains get cited by both ChatGPT and Perplexity on the same prompt.

This means that "AI visibility" is not one channel. It's five separate channels with five separate rules. Optimizing for one engine's citation preferences doesn't automatically improve visibility on another. A strategy that works for Perplexity's retrieval (fresh, structured, externally linked content) is different from what gets a brand embedded in Claude's training data (sustained presence in high-quality sources over time).

The composite score compresses this into a single number for benchmarking. The per-engine breakdown is where you actually plan what to do next.

Reading the report, practically

When we sit down with a report from Be Recommended, we look at it in this order:

1. Composite score for the headline: where does this brand sit relative to the 31 average and the 80+ target?
2. Per-engine scores for the gap analysis: which engines are strong, which are weak?
3. Prompt-level detail for the specific failures: which buying-intent prompts produce zero mentions?
4. Citation quality: mentioned, recommended, or cited with a link?
5. Competitor comparison: is a competitor getting the recommendations this brand is missing?

Steps 3-5 tell you what to actually build or fix. A brand that's invisible on "best X for enterprise teams" but visible on "best X for freelancers" has a positioning gap, not a visibility gap. A brand that's mentioned but never cited with a link has a structured-data or source-attribution problem.

The score is a starting point. The report is the map.

Be Recommended is an AI visibility tool built by Inithouse. It scores how five AI engines recommend your brand and provides a prioritized action plan. No signup required to run a report.