Hiring team finder benchmark
How often can you find the hiring team for a startup job?
In our 2026-11-14 run on 59 live job postings, Hirenote returned at least one hiring contact for 57 and a probable hiring manager for 55. For the 30 startup postings it found someone every time. The two misses were job pages that blocked reading; entering the company and title found four people for each.
Results
| Measure | Before | Now | Startup postings (30) |
|---|---|---|---|
| At least one hiring contact | 42/59 | 57/59 | 30/30 |
| Suggested email at the company domain | 36/59 | 50/59 | 25/30 |
| Probable hiring manager returned | 0/59 | 55/59 | 28/30 |
| Manager title matches the role (strict keyword check) | 0/59 | 17/59 | 9/30 |
| Recruiter title matches the function (strict keyword check) | 0/59 | 23/59 | 12/30 |
Results by job board
| Measure | Greenhouse (17) | Ashby (26) | Lever (5) |
|---|---|---|---|
| At least one hiring contact | 17/17 | 26/26 | 5/5 |
| Suggested email at the company domain | 16/17 | 21/26 | 5/5 |
| Probable hiring manager returned | 17/17 | 24/26 | 5/5 |
| Manager title matches the role (strict keyword check) | 4/17 | 9/26 | 1/5 |
| Recruiter title matches the function (strict keyword check) | 8/17 | 11/26 | 1/5 |
The rest of the sample came from Workday, iCIMS, LinkedIn, Indeed, company career sites and one pasted description.
How we measured it
- We sampled 60 public postings across Greenhouse, Ashby, Lever and other hosts, and recorded whether each sits in our startup audience. One posting expired before the run and is excluded.
- Each posting ran through the same pipeline a user gets, with a fictional resume.
- Every rate divides by all tested postings, including the ones where nobody was found.
Limits of this benchmark
The title checks are strict keyword matches. A manager titled “Director, Engineering” fails a check for “engineering manager”, so these two rows undercount; a hand-checked count is next. Public profiles can be stale, and a matching title does not prove who makes the hiring decision. Suggested emails follow the company’s email format, best match first, and any of them can still bounce. None of this predicts a reply.
Previous runs
We rerun the benchmark when discovery changes and keep each dated result on this page.