How we rate

How Dupple ranks and reviews business and AI software · Last updated September 25, 2026

About · Editorial policy · How we rate · Corrections · For AI agents · Contact

dupple.com has two kinds of ranked content: single-product reviews under /reviews, and "best of" guides under /learn. Some /learn guides include a vendor that paid to be listed, and those guides say so under their title. A rating appears on a dupple.com page only when there is a real basis for it, and the page says what that basis is. No rating is better than an invented one.

Ratings on /reviews pages

A dupple.com review shows one of three things, in this order of priority:

What the page showsWhat it meansHow it is worded
Our editorial scoreScored by us on five criteria (ease of use, feature depth, value for money, integrations, documentation) plus an overall score. 15 reviews currently carry one."We gave [product] 4.3/5"
A user ratingAn aggregate of user ratings, shown with the number of ratings when we have it. It is attributed to users, never presented as our opinion."Users rate [product] 4.5/5 based on 1,204 ratings"
NothingNeither exists, so the page carries no score at all.No score sentence

An older version of the review template derived sub-scores from the product's name. That produced numbers that were not ratings, and it was removed: every score on the site now comes from one of the two sources above.

Prices on reviews

For 92 reviews so far, the pricing section links the vendor's own pricing page with the date it was checked, and where the vendor publishes prices the review states them rather than saying the product does not list pricing. That list grows as reviews are refreshed; a review without the link has not been through that check yet.

Ranked guides on /learn

"Best of" guides rank products by fit for the reader the guide addresses; picks state who they are for, what they cost and their catch. A vendor that paid to be listed can have its position set by that agreement, and the guide names it in a disclosure under the title. Some guides show an aggregated public rating with its review count, taken from the multi-platform aggregate Toolradar maintains (G2, Capterra, Trustpilot and others). When the sample is too thin to mean anything, as for most local LLM tools, the guide shows no rating and says why. When a rating contradicts our own view, we report both and explain the difference rather than hiding the number, as the CI/CD guide does for Jenkins.

Money, ranks and scores

In practice

Questions about a rank, or evidence that one is wrong? Email [email protected] with the page URL. Read the editorial policy for how pages are researched and checked.