Methodology
How our data is sourced and tested
Every field in a WorthPitching lookup shows where it came from and when it was fetched. This page explains each source and shows the results of the test behind the numbers on our homepage.
The 50-company test (October 25, 2026)
We ran fresh lookups on 50 well-known B2B SaaS companies on October 25, 2026, using only free sources (their website, public DNS and public job boards) plus AI checks (Jev, a classification model from TypeSafe). Paid sources (firmographics, Meta ads, search estimates) were not connected, so those fields were empty for everyone.
- 49 of 50
- got 4 or more real fields (of the 6 below)
- 37 of 50
- had hiring data by team from a public job board
- 42 of 50
- had at least one why-now signal fire
- 1.5s median
- per lookup (90% finished within 4.2s)
What counts as a real field
A field counts only when it was filled from a source, never from a guess. The test counted these six:
| Field | Source |
|---|---|
| What they do | Their homepage meta description or headline, kept only when an AI check says it describes the product. |
| Industry | Picked by an AI model from a fixed list, using their website text (kept only above a confidence threshold). |
| Tech stack | Scripts and headers on their homepage, plus public DNS records (MX, SPF, verification TXT). |
| Email provider | Their MX records in public DNS. |
| Hiring | Their public job board (Greenhouse, Lever, Ashby, Workable, Recruitee), found from their careers page. Guessed boards are kept only if the board name matches or an AI check confirms it's the same company. |
| Why now | Rules on the data above, e.g. 3 or more open sales roles. Shown only when a rule fires. |
Results per company
| Company | Real fields (of 6) | Hiring source | Lookup time |
|---|---|---|---|
| stripe.com | 6 | Greenhouse | 2.8s |
| linear.app | 6 | Ashby | 1.3s |
| notion.so | 5 | Ashby | 0.9s |
| figma.com | 6 | Greenhouse | 3.2s |
| vercel.com | 6 | Greenhouse | 0.9s |
| supabase.com | 6 | Ashby | 1.1s |
| posthog.com | 6 | Ashby | 1.6s |
| airtable.com | 6 | Greenhouse | 1.6s |
| asana.com | 6 | Greenhouse | 1.6s |
| hubspot.com | 5 | None found | 3.2s |
| intercom.com | 6 | Greenhouse | 1.8s |
| zapier.com | 6 | Ashby | 0.7s |
| calendly.com | 6 | Greenhouse | 1.3s |
| loom.com | 5 | None found | 0.8s |
| miro.com | 6 | Ashby | 1.3s |
| webflow.com | 6 | Ashby | 1.1s |
| retool.com | 5 | None found | 0.7s |
| segment.com | 6 | Greenhouse | 3.5s |
| mixpanel.com | 6 | Greenhouse | 1.5s |
| amplitude.com | 6 | Ashby | 1.6s |
| pipedrive.com | 6 | Lever | 1.6s |
| close.com | 6 | Ashby | 0.4s |
| apollo.io | 5 | None found | 6.4s |
| lemlist.com | 6 | Ashby | 0.8s |
| instantly.ai | 5 | None found | 0.8s |
| clay.com | 4 | None found | 2s |
| gong.io | 5 | Recruitee | 1.8s |
| outreach.io | 6 | Lever | 1.7s |
| salesloft.com | 6 | Greenhouse | 1.6s |
| attio.com | 6 | Ashby | 0.7s |
| plausible.io | 3 | None found | 0.5s |
| fathom.video | 6 | Ashby | 1.1s |
| cal.com | 4 | None found | 5.9s |
| resend.com | 6 | Ashby | 1.2s |
| clerk.com | 5 | Ashby | 0.8s |
| neon.tech | 4 | None found | 1.5s |
| planetscale.com | 6 | Greenhouse | 0.9s |
| render.com | 6 | Ashby | 1.5s |
| railway.app | 6 | Ashby | 0.8s |
| sentry.io | 6 | Ashby | 1.8s |
| datadoghq.com | 6 | Greenhouse | 1s |
| gitlab.com | 6 | Greenhouse | 1.2s |
| hotjar.com | 6 | Lever | 4.2s |
| typeform.com | 6 | Greenhouse | 5.9s |
| tally.so | 5 | None found | 5.7s |
| pitch.com | 5 | Workable | 0.8s |
| descript.com | 6 | Greenhouse | 1.9s |
| framer.com | 4 | None found | 0.8s |
| deel.com | 5 | None found | 0.9s |
| rippling.com | 5 | None found | 2.2s |
Watchlist signals test
For the watchlist we planted known changes in older snapshots of 20 real companies (removed sales roles, swapped a tool, removed a pricing page) and ran a rescan. All 20 produced the expected signals with evidence and both scan dates, unchanged companies produced none, and a second rescan in the same week produced 0 signals.
Limits
- · The sample is 50 large, well-known SaaS companies. Smaller companies often have less public data.
- · Tech detection sees only what loads on their homepage or shows in DNS, not tools behind a login.
- · Hiring covers companies with a public job board. Jobs posted only elsewhere are missed.
- · Lookup times were measured from Denmark on a single machine; your times will vary.
Questions about a specific field? See the API docs for how sources appear in each response.