AI RECOMMENDATION OBSERVATORY
The YC AI
Visibility Ledger
COHORT INTELLIGENCE
Funding earns attention.
Answers still have to be won.
Recommendation share measures how often a company appears inside a fixed, category-specific prompt set. It is directional, model-sensitive, and more useful as a map of opportunity than a universal league table.
READING THE TABLE Shares are a sampled research signal within each company’s market. Select any row to open its query and citation brief below.
THE CAPITAL / ANSWER DIVIDE
Large companies can still
have small answer surfaces.
AI assistants do not inherit the venture market’s hierarchy. They reconstruct a recommendation from available pages, third-party evidence, current comparisons, and the specificity of the buyer’s need.
COMPANY OPPORTUNITY BRIEFS
Five buyer questions reveal
where the evidence is thin.
Each record shows five high-intent buyer questions, current answer leaders, and the evidence gaps visible in this four-model sample.
Vapi
vapi.ai · 4/4 assistants · 305 returned citation instances
The research already found the work.
We turn each missed query into a managed program across technical access, answer-ready content, and third-party citation acquisition.
Commission Vapi’s 90-day planMake the evidence retrievable
Crawl paths, structured pages, stable entities
Build the missing answer surfaces
Original research, comparisons, decision pages
Create independent corroboration
Digital PR, expert sources, community proof
RESEARCH METHOD
Directional by design.
Specific enough to act on.
This edition compares companies only within relevant buying contexts. It does not equate funding, search traffic, or company size with AI recommendation performance.
Build the cohort
Active YC-backed companies founded or funded from 2020 onward, selected for scale, category relevance, and public buyer interest.
Construct buyer queries
Generic category prompts are combined with workflow, company size, compliance, integration, and budget constraints.
Sample four assistants
The same prompt families are observed across major answer engines and normalized to company entities and product aliases.
Resolve missing surfaces
A query becomes an opportunity when a relevant company is absent or consistently outranked by a less suitable competitor.
Trace citations
Visible source patterns are grouped into owned pages, community proof, editorial authority, reviews, and ecosystem evidence.
Publish a useful preview
Every featured record includes five visible gaps and a citation preview. Deeper verification is completed during a managed research engagement.
Shares are a time-bound sample, not a permanent ranking. No company can pay to change a measured result. Managed-growth work is clearly separated from the public study.
How we executeis a build brief.
Get the private version of your company record.
We’ll verify your highest-value missing queries, expose the full citation graph, and show the exact pages and outreach targets we would execute over 90 days.