Methodology · v1.0
How I test HR & recruiting software
Every score on WiserStaff — on a review page or in a round-up table — comes from the same procedure: I trial or buy the tool on my own account, build one real workflow in it, capture the screens, and score five signals. This page is the full rubric, what disqualifies a top score, and a dated changelog — so you can check how a verdict was produced, and when its facts were last verified.
The 5-Signal Test
Each signal is scored 0, 1 or 2 points, for a total out of 10. A tool is only scored at all once I have run a real workflow in it — until then its review says “Review in progress” instead of a number. The workflows are built the way a 10–50-person team would run them, on the plan tiers that team could actually afford.
| Signal | Points | What it measures |
|---|---|---|
| Core power | 0–2 | Does the tool do its main job well — sourcing and pipeline for an ATS, runs and filings for payroll? Scored on the workflow I actually built, not the feature list. |
| Ease of use | 0–2 | How much of the workflow I could set up without support tickets or documentation spelunking, and how it feels to operate day to day. |
| Value | 0–2 | What the tool delivers against the price I verified in my own account — the in-app number with a capture date, not the marketing page. |
| Fit & integrations | 0–2 | Whether it connects to the rest of a small team’s stack (email, calendar, HRIS, automation layer) without duct tape. |
| Setup & support | 0–2 | Time from signup to a working flow, quality of onboarding, and how support behaves when something breaks. |
What disqualifies a 9 or 10
High scores are capped by evidence, not enthusiasm. A tool cannot score 9 or above unless both of these are true:
- I executed the core workflow end to end in the account myself — not a guided demo, not a sandbox tour.
- The price on the page was verified inside my own account, with a capture date shown next to it.
An 8 can happen two ways, and the review always says which: parts I could not personally exercise (an enterprise quote, a gated module) are scoped out of the verdict rather than scored on faith — or a re-check surfaced a real, dated weakness and points came off for it.
The data behind a review
- Paid or self-registered accounts only. No vendor-provisioned demo environments, no sponsored setups.
- Dated price captures. Prices are read in-app and labeled with the month they were captured. Third-party or advertised figures I could not verify in-account are explicitly labeled as such.
- Screenshots from my own build. The workflow screenshots in a review are from the account I tested, with candidate and employee data kept out.
- No fabricated evidence, ever. A figure I couldn't verify is either left out or shipped with its source label ("advertised", "as reported by…") — never presented as tested. Pages missing their core proof stay unpublished drafts rather than shipping guesses.
Scores move when the evidence moves
A score is a dated judgment, not a lifetime award. When a re-check shows a real weakness — a price change, a feature gate I hit later, support that degraded — the score is revised and the review says why. Re-checks are event-driven (a price change, a vendor announcement, a gate I hit in my own account), on top of a quarterly fact sweep of the live pages.
Every money page dates its trail in the byline: when it was published, when the content last changed (Updated), and when its prices and features were last verified (Facts verified — at publication, then again at each fact re-check, whether or not the text changed).
Independence
Some links on WiserStaff are affiliate links, disclosed above the content on every page that carries them. Which tools get tested is shaped by what small teams ask about — and yes, by which vendors run partner programs. But once a tool is on the bench, the rubric is the same for everyone: Gusto has no affiliate partnership with this site and holds the same 8/10 payroll score as Deel, which does. Any material relationship beyond affiliate links would be disclosed in the review itself. There is no aggregate community rating on this site either: one tester, one rubric, one signed verdict. Full affiliate disclosure →
Changelog
- v1.0 — August 3, 2026. First standalone publication of this methodology page. “Facts verified” dates switched on across every review, round-up, comparison and guide — backfilled from the site’s internal refresh ledger (verification at publication, plus the July 22 sitewide re-check where one applied).
- July 22, 2026. Sitewide fact re-check of live pages: prices, support and feature gates re-verified; Gusto’s score revised against real, dated weaknesses (support, add-on gating) — the review documents the change.
- June 24, 2026. First review published under the 5-Signal rubric.
Independent: not owned by any vendor. Every verdict comes from tools I test with my own accounts and money.