Benchmarks

What B2B homepages tell a buyer, counted

We read every homepage in one B2B category on the same day and count what each one tells a buyer. Every number carries its denominator, and every yes shows the line it came from. Every company has its own row, so anyone can check theirs.

Free to read· Nothing gated· Every chart saves as an image
Published

Read a benchmark

Each one is a single page: the findings, one fix for each, a one-minute self-score, and a free report for your own company.

API tooling · 9 Oct 2026 · 53 homepages

18 of 53 say who they’re for on the first screen.

6 of 53 have a section for the security reviewer. 27 of 53 show no ROI proof and no third-party proof. Kong, Postman, Apollo GraphQL, Mintlify and 49 more. 9 findings, one fix for each.

Read the API tooling benchmark
B2B startups · 9 Oct 2026 · 50 homepages

0 of 50 have a section for the security reviewer.

16 of 50 say who they’re for on the first screen. 20 of 50 show no ROI proof and no third-party proof. Clay, Attio, Linear, PostHog and 46 more. 9 findings, one fix for each.

Read the B2B startup benchmark
Next category

Tell us what to measure next.

Name a B2B category and the homepages that belong in it.

How we measure

The same 11 checks on every page

  • Eleven checks. The four questions a buyer brings (what it is, who it’s for, what it replaces, why it’s better) and seven kinds of proof, each tied to the person on the buying committee it serves.
  • A yes needs a quote. Every yes rests on a line copied word for word from the page, and code confirms the line is there. Nothing is simulated or estimated.
  • One capture date. A real browser opens each homepage at desktop size, closes the cookie banner and scrolls the whole page. “First screen” means the top 900 pixels.
  • Every row is public. Each company has its own row with the line behind every yes. Our own customers and design partners are left out.
  • Corrections. If we misread your page, tell us. We recheck it and fix the row.

The counts show what a page says. To see how each buyer on the committee reads it, see Committee Simulation, and how often its predictions hold up: Accuracy.