How we score
What we score on
Every guide is built on the same four questions, because they are the ones that decide whether a tool works for a given team.
- Depth: does it do the job at clause level, or only summarise?
- Control: can you encode your own positions and standards, or are you accepting the vendor's defaults?
- Where the work happens: in the tools people already use, or in a platform the business has to adopt first.
- Who can actually use it: a qualified lawyer only, or a commercial team without legal training.
How we source prices
We publish a real figure wherever one exists, because a guide that says "contact sales" for every tool has told you nothing. Where a vendor publishes a rate card we quote it. Where they do not, we quote reported figures from marketplace and procurement data and label them as reported, with the range rather than a single number.
Reported figures are directional. Treat them as a way to tell a $200 a month decision from a $200,000 a year one, not as a quote.
The overall score
Every tool carries a score out of 10. It is calculated, not decided. There is no score field in any of our content and the build refuses to publish a page containing one, because a hand-typed overall score is a number anybody can nudge until the ranking matches the one they already wanted.
It is built from three things, weighted like this:
- Position, 90%: where the tool sits in that guide. The order reflects our judgement of which tool best does the job the page is about.
- Scope, 5%: how much of the job the tool does, from a single step to a whole lifecycle.
- Price transparency, 5%: Published scores 5, Reported 4, None 3.
Every tool is scored out of 10. The fit scale measures fit rather than quality: a tool that scores 1 for three of our four buyers is narrow, not bad, and every tool on this site is a real product somebody is using today. A lower score means the tool serves fewer of the buyers we score, not that it is failing.
We changed how this works on 3 September 2026. Until then every tool carried one score across the whole site, worked out from how well it suited four different kinds of buyer. The problem was that a single number cannot be right on two pages at once: a tool can be an excellent choice for an in-house team and the wrong choice for a law firm, and one averaged figure was misleading on both. Scores are now specific to the guide you are reading.
Scope and price transparency are weighted small on purpose. They separate two tools that would otherwise land on the same number, and they are never large enough to push a tool above the one listed before it.
Two things the score is not. It is not the ranking: guides are ordered by fit for the buyer named in the title, so a tool with a lower score can and does come first on a page whose buyer it suits best. And it is not a per-page judgement: it is derived from fit scores held in a single file, so a tool scores identically on every guide it appears on and cannot be tuned to suit an individual page.
What we do not score on
Integration counts, security badges and model names. Nobody chooses a contract tool because it lists forty integrations, and every vendor claims the same three certifications. Where security genuinely differs we say so in the tool's entry rather than awarding points for it.
We also do not score on feature totals. A tool that does one job properly beats a tool that does nine badly, and a feature count hides that.
Ranking
Guides are ordered by fit for the buyer named in the title, not by the overall score. The two often disagree, and where they do the ranking is the considered answer and the score is a summary. A tool ranked fifth is often the right answer for someone, which is why every entry carries a "best for" line. Where two tools suit the same buyer, the one with the clearer path to getting started ranks higher.
Updates
Each guide shows the date it was last checked. We re-verify prices and product claims on a monthly cycle, and correct in place. We do not re-date a page we have not actually rechecked.