How we research
Every review is desk research. We read vendor documentation and pricing pages, review platform listings such as G2 and Capterra, vendor case studies and independent coverage, and we compare what each source says. We do not use the products, and no page claims that anyone used, tested or trialled a product.
Every number on a page carries a named source type. Vendor-sourced means the vendor published it. Third-party reported means an independent publisher stated it. Review platform means it comes from a listing such as G2 or Capterra. Self-reported means a user or provider stated it without a way to check it. When sources disagree, a page shows both values and says which source gave which. A claim that cannot be verified is left out of the page.
The rating scale
Each criterion is scored out of 10 with one decimal. The overall score is the weighted average of the six criteria, rounded to one decimal and computed in code, with these weights.
| Criterion | Weight |
|---|---|
| Proof of results | 30% |
| Output quality | 25% |
| Ease of use | 15% |
| Pricing and value | 15% |
| Customer support | 10% |
| Security and compliance | 5% |
The same overall score appears on a tool's review, in the roundup that features it and on any comparison or alternatives page that cites it.
The six criteria
Proof of results
Weight 30%. This criterion rests on review platform volume and ratings with the platform named, named customers and documented customer outcomes. Outcomes that a vendor publishes are labelled vendor-sourced and are not averaged with independent data.
Output quality
Weight 25%. This criterion rests on documented capability and aggregated user reports about what the product does. The label applies to every category, including categories where the product's output is not content. AI features are judged under this criterion.
Ease of use
Weight 15%. This criterion rests on onboarding and interface comments from several reviewers on review platforms and in independent coverage.
Pricing and value
Weight 15%. This criterion rests on whether prices are published, how far the entry price sits from the total cost, per-seat steps, required modules and add-ons, and conflicting reports. Opacity and sharp price steps lower the score even when the product is strong.
Customer support
Weight 10%. This criterion rests on support options by plan tier and on the recurring themes in review platform feedback.
Security and compliance
Weight 5%. This criterion rests on published security certifications such as SOC 2 or ISO 27001, data export, single sign-on and data protection terms. A review mentions it only where something matters to a buyer, such as single sign-on offered only on a sales-quoted tier.
How rankings and ties work
Roundups list tools by overall score, highest first. Tools with equal overall scores are listed by name.
What we do not do
We do not use or trial the products we review. We do not state a claim about a product that no source supports. We do not average vendor claims with independent data.