Every product on WhatsTheStack is scored with the same public framework, so an 8.5 for a CRM means the same thing as an 8.5 for a password manager: an excellent choice for most small businesses in its category, with trade-offs worth understanding. This page is the short version; the full rubric, weights and rules are on our methodology page at /methodology.
How a review is researched
- Define the buyer: we write down who the product is for (team size, budget, technical skill) before scoring it.
- Gather evidence: vendor documentation, help centre and security pages, changelogs and the public pricing page, recorded with the date checked.
- Score six criteria: Features & fit, Ease of use, Value for money, Support & reliability, Security & privacy, and Integrations & ecosystem — each against written 1–10 anchor definitions.
- Calculate the overall score: a weighted average using fixed weights for the product's pillar. The calculation is done by code; nobody enters an overall score by hand.
- Record confidence and dates: each review shows its evidence, a confidence level, when it was last verified and when it is next due for re-review.
- Log changes: any change to a score is recorded with a date and reason in that review's history.
Research vs hands-on testing
Each review carries a status badge. 'Researched' means the review is based on desk research — vendor documentation, dated public pricing, product demos and changelogs — not on our own extended use; those reviews carry Medium confidence. 'Hands-on' means an editor has used the product for real tasks against our test script; those reviews carry High confidence. At launch, all reviews are 'Researched', and we upgrade them as hands-on testing is completed.
The six criteria
- Features & fit: does it do the jobs a small business buys it for, on a realistic plan?
- Ease of use: how quickly can a non-specialist team get value, and how much admin does it need?
- Value for money: what you get at 5, 10 and 25 seats, including tier jumps, add-ons and renewals.
- Support & reliability: support channels and speed, documentation, uptime and status transparency.
- Security & privacy: MFA/SSO, encryption, certifications, data-training and retention policies, incident history.
- Integrations & ecosystem: native integrations, API, automation support and data portability.
Weights differ by pillar — for example, Security & privacy carries 30% for security products but 10% for general business software. The weights table is published on the methodology page.
What the scores mean
- 9.0+ — Outstanding: best in class for most buyers.
- 8.0–8.9 — Excellent: a strong choice with minor trade-offs.
- 7.0–7.9 — Good: right for specific needs.
- 6.0–6.9 — Fair: consider alternatives first.
- Below 6.0 — Weak: significant gaps.
Keeping reviews current
AI products are re-checked at least every 3 months and everything else at least every 6 months, plus whenever pricing or plans change. Every review shows a 'Last verified' date and the next scheduled check. If you spot something out of date, please tell us.
