How products get their score.
Every product we rank has a score out of 100 and a set of sub-scores. This page explains where those numbers come from, what they mean, and what they are not.
Methodology v2.0, October 2026
Four steps from launch to listing.
There is no testing lab behind these scores. They are research-based judgements, drafted with AI and checked against sources.
Find the product
New products mostly come from Product Hunt launches in each category. Established products were added from public market research.
Research public sources
We collect what is publicly available: the vendor's site, documentation, pricing, the launch listing, and public registries where relevant.
AI-assisted scoring
An AI model drafts the review, the overall score and the sub-scores from that research, using the dimensions below. Automated checks clean up formatting and common AI writing problems.
Publish and correct
The product goes live in its category, ranked by score. When we learn a fact is wrong, from our own checks or from a reader or vendor, we fix it and revise the score if needed.
The dimensions shown on each review.
Each review shows sub-scores out of 100 for the dimensions that fit its category. The overall score is a judgement across them, not a fixed weighted formula.
Output quality
How well the product does its core job.
Ease of use
How quickly a new user gets to a useful result.
Pricing value
What you get for the price, and how clear the pricing is.
Integrations
How well it connects to the tools around it.
Speed
Responsiveness and performance, where public information exists.
Support
Documentation and the support options on offer.
Trust and reliability
Track record, security posture and stability.
Ecosystem
Community, templates, partners and momentum.
What the bands mean.
Best in class
Leads its category with no significant weakness we could find.
Excellent
Strong across the board, with minor gaps.
Good
A solid choice with real trade-offs.
Acceptable
Fits narrow use cases; the gaps matter.
Poor
Significant problems. Look elsewhere first.
Scores compare products within one category. A 75 in hosting and a 75 in design tools are not directly comparable. Star ratings on reviews are our editorial rating, not an average of user reviews.
What our scores are not.
Not lab results
We do not benchmark products ourselves. Performance figures come from public sources and are attributed.
Not user ratings
We do not collect user reviews. Any rating shown is ours.
Not for sale
Sponsors cannot buy a score or a position. Paid mentions are labelled Sponsored.
Not always current
Products change after we review them. Each review shows when it was published or last updated.
Tell us, with a source.
Email corrections@gaxonline.com with the page URL, what you think is wrong, and a link to a public source that shows it. We check the claim against that source and update the page when it holds up.
v2.0, October 2026: rewritten to describe the process we actually use. The earlier text described testing, interviews and a quarterly re-score that were never part of how the site worked.