These B2B Brands Are Completely Invisible to Every AI Search Engine

FogTrail tracked 25 B2B SaaS brands across 5 AI search engines (ChatGPT, Perplexity, Gemini, Grok, Claude), 20 queries, and 3 weekly waves in March 2026, producing 300 engine-query pairs per brand. Two startups scored zero citations across every single pair. Height, a project management tool, was invisible across all 300 engine-query pairs: zero citations, zero mentions, from any engine, on any query, in any wave. Attio, a CRM platform, was invisible across 200 pairs tracked over 2 waves. When buyers asked AI engines about their categories, these brands simply did not exist.

This is not a ranking problem or a visibility dip. It is total absence. And the data suggests it is a structural condition that can affect any startup without a deliberate AI visibility strategy.

The Bottom Line

8% of tracked B2B brands are completely invisible across all AI engines, all queries, and all waves. Both invisible brands are startups.
Enterprise brands average 16.8 mentions per brand and appear on 5.0 of 5 engines. Startups average 6.6 mentions and appear on 2.9 engines.
Invisibility is not inevitable for startups. PostHog, an open source analytics startup, grew its citations from 2 to 5 across 3 waves and was cited by all 5 engines by Wave 3.

What Zero Out of 300 Actually Means

Height was not cited by ChatGPT. Not by Perplexity. Not by Gemini, Grok, or Claude. Not on any of the 20 queries FogTrail tracked across the project management, analytics, CRM, developer tools, and email marketing categories. Not in Wave 1, not in Wave 2, not in Wave 3. Three hundred chances to appear, and the result was silence every time.

Attio followed the same pattern across 200 engine-query pairs in the CRM category. Two waves of queries, five engines each, and not a single mention.

To put this in perspective: when a buyer asks ChatGPT "what's the best CRM for startups" or asks Perplexity "project management tools for small teams," these brands are not in the running. They are not ranked low. They are not mentioned with caveats. They are absent from the conversation entirely.

This matters because AI search is increasingly where B2B buying research begins. A brand that does not appear in any AI engine's response to any relevant query has, for a growing segment of buyers, no presence at all.

The Enterprise vs. Startup Visibility Gap

Enterprise brands in FogTrail's study averaged 16.8 mentions per brand across the 3-wave study and appeared on an average of 5.0 out of 5 engines. Startups averaged 6.6 mentions and appeared on only 2.9 engines. The gap is roughly 2.5x in raw mention volume and nearly double in engine coverage.

This is not surprising. Established brands benefit from years of accumulated web presence: hundreds of indexed pages, Wikipedia entries, press coverage, G2 and Capterra listings, community discussions on Reddit and Hacker News, and integrations documented across partner ecosystems. AI search engines use retrieval-augmented generation (RAG) to pull answers from indexed content, and enterprise brands simply have more retrievable surface area.

But the gap between "fewer citations" and "zero citations" is qualitatively different. A startup with 2 or 3 mentions is in the conversation. A startup with zero is not part of the information AI engines draw from at all.

PostHog Proves Invisibility Is Not Inevitable

PostHog, an open source product analytics platform, started FogTrail's study with just 2 citations in Wave 1. By Wave 2, that number grew to 3. By Wave 3, PostHog had 5 citations and was being cited by all 5 engines. That trajectory, from near-invisible to broad engine coverage in 3 weeks, shows that startups can build AI search presence.

What separates PostHog from the invisible brands is not budget or company size. PostHog has extensive public documentation, an active blog covering product analytics topics in depth, a strong GitHub presence, regular mentions on Hacker News and developer forums, and third-party reviews on comparison sites. In other words, PostHog has the exact set of signals that AI retrieval systems look for when deciding what to cite.

Height and Attio, by contrast, lack the volume of independently retrievable content that would give any engine enough material to reference. The product may be excellent. The retrieval system does not know that.

Why Some Brands Have No Retrievable Footprint

AI engines cannot cite what they cannot find. A brand becomes invisible when it fails to produce signals across four categories that retrieval systems depend on.

No third-party coverage. If the only source on the internet saying your product exists is your own website, AI engines treat your claims as unverifiable. ChatGPT in particular weighs independent corroboration heavily. Without reviews on G2, mentions on comparison sites, or press coverage, the retrieval system has no external confirmation to work with.

No community discussion. Grok cites Reddit 13x more than Claude, Perplexity, and Gemini combined. Perplexity pulls heavily from forum threads and Q&A sites. If nobody is discussing your product in public forums, you are missing an entire channel of retrievable mentions.

No content depth. A landing page and a docs site are not enough. AI engines assess whether a domain has topical authority. A brand with 50 articles covering use cases, comparisons, how-to guides, and category analysis signals depth. A brand with a homepage and a pricing page signals nothing. The retrieval system needs candidate passages to evaluate, and a thin content library offers none.

No structured extractable claims. Even if content exists, AI engines need clean, self-contained passages that directly answer a query. Buried claims, vague descriptions, and marketing-heavy copy do not produce the kind of passages that retrieval systems select for citation.

The Compounding Problem

Brands that are invisible today face a compounding disadvantage. AI engines are nondeterministic, meaning responses vary between identical queries, but they are not random. Brands that appear in early results build a feedback loop: citations generate traffic, traffic generates engagement signals, engagement signals feed back into the retrieval systems that select future citations.

A brand starting from zero has no momentum to build on. Every week of invisibility is a week where competitors accumulate the signals that make them more likely to be cited next time. FogTrail's wave data shows this pattern clearly: brands that were cited in Wave 1 tended to maintain or grow their citations in Waves 2 and 3. Brands that started at zero stayed at zero.

This is not a problem that resolves on its own. Without deliberate intervention, building the content and third-party presence that retrieval systems need, invisible brands tend to remain invisible.

What Invisible Brands Can Do About It

The path from zero to cited requires building retrievable surface area across the four signal categories described above. Based on the patterns FogTrail observed in brands that do get cited, the highest-impact actions are:

Build topical content depth. Create 30 to 50 articles covering your category from multiple angles: comparisons, use cases, how-to guides, and data-driven analysis. Each article should contain clean, self-contained passages that answer specific queries a buyer might ask an AI engine.
Get on review and comparison sites. G2, Capterra, and category-specific directories are heavily indexed by AI retrieval systems. A presence on these platforms creates the third-party corroboration that engines require.
Generate community discussion. Participate in relevant Reddit threads, Hacker News discussions, and industry forums. Organic mentions in community contexts are a retrieval signal that several engines weight heavily.
Publish structured, extractable content. Every page on your domain should have clear headers, direct claims in the opening sentences of each section, and enough specificity that a retrieval system can extract a useful passage without needing surrounding context.
Monitor across all engines. A brand might appear on Gemini but not ChatGPT, or on Claude but not Perplexity. Each engine has different retrieval behavior, different source preferences, and different citation patterns. Single-engine monitoring misses the full picture.

The startups in FogTrail's study that achieved and grew their citations share one trait: they had invested in building the kind of web presence that retrieval systems can draw from. The invisible brands had not. As of April 2026, the gap between those two groups is widening.

Frequently Asked Questions

How do I know if my brand is invisible to AI search engines?

Search for your brand's category queries across all 5 major AI engines: ChatGPT, Perplexity, Gemini, Grok, and Claude. If your brand does not appear in any response across multiple queries, you are likely invisible. Single-query checks are insufficient because AI responses are nondeterministic. You need to test across multiple queries and multiple engines to confirm a pattern. An AEO platform like FogTrail ($499/mo) automates this across 100 queries and 5 engines with 48-hour refresh cycles.

Can a brand with a good product still be invisible to AI search?

Yes. AI search engines do not evaluate product quality directly. They cite based on what their retrieval systems can find and extract from indexed web content. A brand with an excellent product but no third-party reviews, no content library, and no community discussion will be invisible regardless of product merit. Height and Attio are both well-regarded products that scored zero citations across hundreds of engine-query pairs.

How long does it take to go from invisible to cited by AI engines?

Based on FogTrail's 3-wave study, brands with existing content depth can see citation growth within weeks. PostHog grew from 2 to 5 citations across 3 weekly waves. Brands starting from true zero, with no content library and no third-party presence, should expect a longer timeline of 2 to 4 months to build sufficient retrievable surface area. The key variable is how quickly you can create the content and third-party signals that retrieval systems need.

Do all AI search engines treat invisible brands the same way?

No. Each engine has different retrieval behavior and source preferences. ChatGPT links to brand websites in 24% of citations versus 2% for Grok. Grok cites Reddit 13x more than Claude, Perplexity, and Gemini combined. A brand might be invisible on one engine but cited on another, depending on where its existing content and mentions are indexed. Total invisibility across all 5 engines, like Height and Attio experienced, indicates a fundamental absence of retrievable signals.

Is AI search invisibility permanent?

No. Invisibility is a structural condition, not a permanent state. It results from the absence of content and third-party signals that retrieval systems need. Brands that build topical content depth, establish review site presence, and generate community discussion can move from zero to cited. The compounding nature of citations means that starting sooner produces better results, but no brand is permanently locked out.