Editorial · Methodology
A transparent look at how SoftFinders evaluates software: evidence gathered from official vendor sources, checked independently, and turned into a score by fixed rules.
What it means
We’ve researched 800+ AI software tools across a wide range of categories, which gives us a broad view of how products differ in capability, usability, pricing, and support. Every SF Score is built the same way: evidence from the vendor’s own sources, assessed against 20 published factors, then calculated by fixed rules rather than opinion.
What SF Score measures
Every product is assessed against the same 20 factors, grouped into four dimensions and weighted by what matters most to buyers.
What the product actually does
Whether the software does the core job it is sold for, and how far its capabilities go.
What it takes to get value
The effort involved in starting, running, and maintaining the product day to day.
What you get for the price
How clear the pricing is, and how much of the product you can genuinely use at each tier.
What happens when you need help
The help a buyer can actually reach, and what the vendor commits to publicly.
The process
We research first party vendor sources: official product and feature information, pricing, and, where a vendor publishes them, documentation and support resources. Evidence used in scoring must be traceable to an official vendor source our research system has verified, and the first assessment is always checked independently.
From evidence to number
Research and calculation are deliberately kept apart, so a score can always be traced back to what a vendor actually published.
AI research reads official vendor sources, pulls out the relevant evidence, and classifies it against our published criteria. It never invents or calculates the number.
Once the factor findings are settled, the SF Score is calculated from them by fixed rules and weights: Feature Quality 30%, Ease of Use 25%, Value for Money 25%, and Support 20%. The same findings always produce the same score.
If reliable evidence cannot establish a factor, we mark it not established and move on. A vendor saying nothing about something is not treated as a failure, and missing information never becomes a zero.
When too little is established, our methodology decides whether a dimension or the overall score can be published at all. We would rather leave a score unavailable than publish one built on weak evidence.
Important scoring decisions are independently verified rather than resting on a single AI judgment. Where assessments disagree, we review the disputed factors again instead of averaging conflicting answers.
Unusually high or low results receive additional scrutiny before publication rather than being accepted automatically.
Independence
An SF Score reflects what a vendor’s own sources establish, and nothing else. Commercial relationships and market presence have no effect on it. Sponsored placements can appear on SoftFinders, but they are clearly labeled Sponsored and kept separate from scoring.
Always current
Scores represent the evidence available at the time of assessment and may be updated as products, pricing, documentation, or support offerings change.
Revisited where practical as products change
Reassessed when a product changes meaningfully
Updated when plans, limits, or costs move
FAQ
No. AI research helps us read official vendor sources and match what they say against our published criteria. It never invents or outputs the number. Once the factor findings are settled, the score is calculated by fixed rules and weights, so the same evidence always produces the same result.
Explore researched software lists, compare buyer-fit signals, and choose tools with clearer context.
