The Fit Score Methodology
A single star rating cannot tell you whether a tool is right for your specific job. Fit Score replaces that with five scored dimensions, judged against a real, already-documented use case, using a published rubric anyone can check us against.
Claude is not simply "good" or "bad." It might be an excellent fit for legal research and a poor fit for generating short-form video. A single aggregate rating flattens that difference away. Fit Score is our attempt to keep it: instead of one number for a whole tool, we score how well a tool fits one specific, real use case at a time.
The Rubric
Every Fit Score is built from five axes, each scored from 1 to 5. We show all five, plus a simple average, rather than folding everything into one opaque number.
| Axis | What a 5 looks like |
|---|---|
| Capability | The tool has the features and depth this use case actually needs, confirmed against vendor documentation and our own hands-on notes for that tool. |
| Reliability | Nothing in the tool's documented track record, maturity, or known limitations (as noted in our pros and cons for it) points to this use case being a weak spot. |
| Value | The published price of the plan this use case requires is proportionate to what you get for it. This is not a claim about which tool is cheapest overall, only whether the price fits what this specific job needs. |
| Ease | Little to no setup, configuration, or learning curve stands between signing up and getting this specific use case working. |
| Compatibility | The tool fits cleanly into a normal workflow for this use case: documented integrations, API availability, and compatible input or output formats. |
A 1 on any axis means the opposite: the tool documentedly lacks what this specific use case needs on that dimension. The overall score shown for a tool and use case pairing is a plain average of the five axis scores, rounded to one decimal, e.g. 4.2 out of 5. We deliberately kept the scale coarse (1 to 5 per axis, not 1 to 100) so the result reads as the editorial judgment it is, not as false precision.
How a Score Gets Made
A Fit Score is never assigned to a tool as a whole. It is assigned to one tool against one specific, already-documented use case, pulled from that tool's own existing bestFor or use case list on this site, the same fields every tool page already publishes. We do not invent new use cases just to have something to score.
Every axis score is grounded in evidence that already exists on this site or in the vendor's own published materials: the tool's pricing page, its documented features and integrations, and our own editorial verdict and pros and cons for that tool, built the same way as every other tool page here. Nothing is invented to make a score look more official or more precise than the underlying evidence supports. This is the same sourcing discipline behind the trust and pricing data already published on every tool page.
Current Scope: a Narrow Pilot
Fit Score is not yet applied across the full catalog of 389+ tools. We are starting deliberately small: an initial batch of roughly 30 to 40 of our most featured, highest-traffic tools, each scored against their top two or three already-published use cases. Full-catalog coverage is a separate decision for later, made only if this pilot proves useful, not a commitment made here.
Scores appear directly on the relevant tool pages as they are published. If a tool does not show a Fit Score yet, it simply has not been reached in the pilot batch.
What This Is Not
- Not a certification or compliance score. We do not audit or certify a tool's security, privacy, or regulatory posture; see each tool's own facts (privacy policy and trust page links) for that.
- Not a lab-measured performance benchmark. We do not run timed tests, load tests, or output-quality benchmarks. Every score reflects documented evidence and our own editorial judgment, not measured results.
- Not paid placement. No tool can pay for a higher Fit Score, and a tool's advertising or affiliate relationship with this site, where one exists, never factors into a score.
- Not comprehensive, yet. During the pilot, most tools and most use cases will not have a Fit Score at all. Absence of a score is not a judgment; it means that pairing has not been scored.
Keeping Scores Current
A score is revisited whenever the underlying evidence changes: a pricing update, a revised editorial verdict, or a tool being marked discontinued all trigger a re-check, the same way pricing and trust data get re-verified elsewhere on this site. We also run periodic re-scoring passes as part of routine catalog maintenance. The "Was this useful?" feedback prompts already live across tool, comparison, and workflow pages collect anonymous signal that may inform future scoring reviews, alongside the vendor and editorial evidence the rubric already relies on.
Methodology version 1.0, published September 2026.