How we review and score
Every app on saasreview is checked the same way, against the same scorecard, so the score means the same thing whether you have one user or one million. Here is exactly how it works, so you and the AI tools you ask can trust it.
The ten things we score
Each app is scored 0 to 10 on every dimension below. The scorecard is product-first: it rewards what the maker actually built and controls, so the weights are not equal.
Problem it solves
Is there a real, painful problem here, and does the product actually solve it?
UX & alignment
Can a first-time visitor do the main thing without getting lost or stuck?
Use case
Is it clear who this is for and the exact job it does for them?
Performance & reliability
Does it load fast and stay up? Measured (LCP, CLS, errors), not guessed.
Docs & policies
Are there real docs, terms, and a privacy policy, the things a shipped product carries?
Innovation
Is there a genuine edge, or table stakes with a fresh coat of paint?
Look & feel
Does it look credible and intentional, or unfinished?
Discoverability (AEO)
Can people and AI assistants actually find, read, and cite it?
Demand
Is there evidence people want this yet? Weighted low so a new app is not punished for being new.
Trust & credibility
Does it earn trust honestly, without fake social proof?
how the score is calculated
The headline score out of 10 is a weighted roll-up of the ten dimensions above, never a number anyone types in by hand. You can always reproduce it from the scorecard on the review. Only dimensions we could actually measure are counted, and the weights are renormalized over exactly those, so a missing dimension never quietly counts as a zero or drags the rest down.
Agents do the analysis, a person checks it
You submit a link
You give us the live URL (and an optional test login so the agents can go behind your sign-in). No account needed.
Agents remove the bias
Several AI agents run the exact same checks on every app against the scorecard below: what works, what is broken, what is exposed, how fast it loads (measured, not guessed), and whether people and AI tools can find it. Same standard for everyone, so there is no flattery and no friends-and-family bias.
A real person checks the facts
Before anything is published, a person at saasreview reads the findings and pulls anything that is not honest or does not hold up. We will pass on publishing a review we cannot stand behind.
We publish, and re-audit
The review goes live with a score, a verdict, and a plain list of what to fix. When you ship changes we can re-audit and publish the full before-and-after, so the history is always visible.
Independent, and we say when it is paid
The same scorecard runs on every app, so paying changes the work we do and the listing you get, never the score. The score is ours to decide.
Every review carries a clear label. Reviews someone paid for are marked as such. Reviews we start ourselves are marked independent. Bughunt security findings are private and never published. We would rather pass on a review than publish one we cannot stand behind.
When you ship fixes, we can re-audit and publish the full before-and-after on your review, so nothing is edited silently and the history stays visible.
Questions about the score
Who decides the score, a person or AI? ▾
The agents produce the per-dimension scores from the same checks every app gets, which is what keeps them consistent and unbiased. The headline score is then calculated from those, not picked by anyone. A real person checks the findings for honesty before publishing, but does not hand-set the number.
Is the score just the average of the ten dimensions? ▾
No. It is a weighted roll-up, so the things a maker actually built and controls (the problem fit, the experience, performance) count for more than signals that mostly reflect having users already (demand, social proof). Only the dimensions we could actually measure are counted, and the weights are renormalized over those, so a missing dimension never quietly counts as a zero.
Are reviews independent if someone paid for one? ▾
Yes. You pay for the review work and the listing, never for a good score. The score is ours to decide, the same scorecard runs on every app, and every paid review is clearly labeled. Reviews we start ourselves are labeled independent. We will pass on publishing a review that is not honest.
Why weight a brand-new app fairly? ▾
A genuinely good but pre-traction app should not be punished just for being new. So demand and trust signals that mostly reflect existing users carry the least weight, and problem fit, experience, and performance carry the most.
Get the same honest read on your SaaS
A real score out of 10, a plain list of what to fix, and a badge you can show off. From $5.