The AI visibility ceiling: why even top B2B brands miss 20-56% of buyer queries
We measured 270 B2B software brands across four web-grounded AI engines. The best-performing brand in customer support achieved an 80% mention rate. In project management, the ceiling was 52%. In HR software, 44%. No brand in any category approached 100%. That gap is structural, not a content problem.
What the ceiling looks like by category
Across 270 brands run against 10 buyer-intent prompts each, here is how the top of each category performed:
| Category | Max mention rate | P90 | P75 | Median |
|---|---|---|---|---|
| Customer support software | 80% | 39% | 33% | 27% |
| Project management software | 52% | 40% | 30% | 20% |
| HR software | 44% | 32% | 29% | 20% |
The 80% ceiling in customer support means the top brand in that category was absent from 2 out of every 10 buyer-intent queries -- even though it was the single most visible brand studied. In project management, the top brand missed nearly half. In HR software, the top brand missed more than half.
Why 100% is not achievable
The gap is not caused by bad content or weak brand authority. Three structural factors create a ceiling for every brand in every category.
AI engines rotate their source retrieval per query. The pages a search-grounded AI engine reads when answering "best project management software for remote teams" are not the same pages it reads for "project management tools for construction companies." Each query variation pulls a different set of sources, and no brand appears as the top source across every possible source set for every possible query framing. A brand that dominates the sources for one prompt type is not retrievable from the sources that surface for adjacent prompt types.
Query diversity spans niches a single brand cannot own. A buyer-intent prompt panel for "project management software" includes queries about remote teams, construction projects, agency workflows, agile development, and budget tracking. Each sub-niche surfaces different recommended brands. A PM tool built for software teams may score 50% on dev-focused queries and 0% on construction queries. Its overall mention rate is the average across all prompt types, not the maximum in any one.
Multi-engine diversity means different source sets. ChatGPT, Gemini, Perplexity, and Claude retrieve pages through different indexing and ranking mechanisms. A brand well-sourced on Perplexity's index may not appear in the pages Gemini retrieves for the same query. Scoring across all four engines blends what each engine independently surfaces, so a brand must appear in multiple engines' source sets -- not just one -- to achieve a high overall score. The brands in this study with the highest mention rates tend to appear consistently across all four engines, not just one.
What this means for strategy
If 100% is not achievable, the question is where in the distribution to aim. The data suggests a clear answer: the top quartile, not the maximum.
| Target mention rate | Field position (all categories) | What it signals |
|---|---|---|
| Under 20% | Bottom half | Absent from most buyer-intent queries. Competitors are filling the gap. |
| 20-30% | 50th-67th percentile | Median range. Present in some queries, invisible in most. |
| 30-40% | 67th-91st percentile | Top third. Consistently appearing across the major query types. |
| 40%+ | Top 9% | Category leadership position. Saturating the sources the engines retrieve. |
The jump from the 20% median to 30% (top third) requires consistent coverage on the three or four editorial sources that each engine retrieves most frequently for buyer-intent queries in your category. That is a targeted task. The jump from 30% to 40%+ (top 9%) requires coverage across multiple query framings and multiple engines -- a harder but still bounded problem.
Chasing 100% is not a strategy; it is a category impossibility. Chasing 40% in your specific category is a concrete goal with a specific set of sources to target and a measurable way to verify progress.
The ceiling differs by category for a reason
Customer support software has a higher ceiling (80%) than project management software (52%) or HR software (44%). This is not random. Customer support is a more concentrated category: a smaller set of brands (Freshdesk, Zendesk, Intercom, Help Scout) capture the majority of all AI recommendations, and the top brands are strongly represented across the main editorial sources. A brand that builds strong coverage in that narrow source set can achieve a high score because the source concentration works in its favor.
Project management software is more fragmented by use case: construction, software development, agency work, product management, and event planning each pull from different comparison pages and editorial roundups. No brand dominates across all of them, so the ceiling is lower. HR software sits in between.
Your category's ceiling is the realistic upper bound on what consistent investment can achieve. Knowing it helps you set a goal that is aggressive but not impossible.
The gap the ceiling exposes
Even a brand at the ceiling is absent from a significant fraction of buyer queries. At 80%, 1 in 5 potential buyers asking AI about your category does not hear your name. At 52%, nearly half do not. At 44%, more than half do not.
Those unanswered queries are not silent. The AI engine recommends a competitor instead. The distribution of who fills that space is narrow: in this data, 2-4 brands absorb nearly all the buyer attention in every absent brand's place. If your brand is below the ceiling, you are sending buyers to one of those few brands every time you are missing from a query.
The ceiling is not a reason to stop investing. It is a reason to invest precisely: understand which query types you are missing, identify which sources those queries retrieve, and build presence there. That is a smaller set of actions than it sounds, and the payoff is measurable.
We measure your brand's mention rate across four AI engines, map which query types you are missing, and identify the specific sources your top competitors are using to fill that gap. The audit tells you where your ceiling is and what it would take to reach it.
Get the AI Visibility Audit, $490Methodology note
270 B2B software brands were measured in June 2026 using live, search-grounded AI endpoints: ChatGPT (OpenAI Responses API with web search), Gemini (with grounding), Perplexity (Sonar), and Claude (with web search). Each brand received 10 buyer-intent prompts generated for its specific category and geography. Mention scoring used automated word-boundary matching against the brand name. All results reflect organic buyer-intent queries only. Percentiles cited (P75, P90) reflect the full 270-brand distribution.