Editorial policy

How we test and score

Every score on this site comes out of the same rubric. Here it is in full.

Last updated August 2026

The five axes

Each tool is scored 0–10 on five criteria, then weighted into a single number.

AxisWeightWhat it measures
Output quality34%Whether what comes out is publishable without a rewrite. Judged blind against the same briefs.
Features22%Depth of what the tool does versus what its category demands, not raw feature count.
Value20%What the realistic monthly bill buys compared with direct competitors at the same tier.
Ease of use16%Time from signup to first useful output, and how often we hit a wall.
Support8%Response times on real tickets we file, plus documentation quality.

What we do not do

  • We do not accept payment for placement, ranking, or a review at all.
  • We do not send drafts to vendors for approval before publishing.
  • We do not use free vendor-provided accounts where a paid plan is what readers would buy.
  • We do not leave a score up once it is stale — reviews are re-tested at least twice a year.

Affiliate relationships

Most tools we cover run affiliate programs and we join them. Commission rates vary widely between vendors, which is exactly why rate has no input into the score. Tools with no program at all are reviewed and ranked identically — several currently outrank tools that pay us.

Freshness

Pricing is re-checked monthly and stamped on each review. If you see a price on this site that no longer matches the vendor's page, that is a bug and we want to hear about it.

Use of AI in our own content

We test AI tools; we do not let them write the reviews. AI assists with transcript cleanup and first-pass data extraction from pricing pages. Every judgement, score and sentence of analysis is written and checked by a human who used the product.