How we rank

Every alternatives page starts from one product, the one being replaced. Its documented capabilities are listed in a grid of 15 to 24 features, each weighted 3 (it defines the product), 2 (common use) or 1 (secondary). Price is not in the grid.

Each alternative gets one mark per feature: 1 when the vendor documents a native equivalent, 0.5 when the equivalent is partial (a documented limit, a paid add-on, a third-party integration), 0 when it is absent or could not be verified. The similarity is the weighted sum of the marks divided by the sum of the weights, rounded to a whole percentage.

The ten entries are the ten highest results among at least twelve candidates, in descending order. A tie goes to the tool whose entry price is closest to the starting product, then to alphabetical order. The percentage shown on a page is computed from the grid and the marks the page carries; it is never typed by hand.

A category top 10 has no starting product. The grid is the reference grid of the category: the capabilities documented by at least three vendors of that category. The same calculation gives a score out of 100, which reads as coverage of the expected capabilities, not as a quality rating.

Every price, limit and feature is read on the vendor's own pages (pricing, documentation, help centre, terms) and recorded with its address and the exact wording. Comparison sites and third-party articles are not sources. A fact that could not be confirmed on a vendor page is left out, or stated as not verified with the date of the check.

A page is written by one session and verified by another, which reopens every price and a sample of the other facts before publication. Prices are re-read every 30 days and feature marks every 180 days; the date under each title is the date of the last reading.

Reasons-for-leaving pages show a percentage only when a published source gives one. Otherwise each reason is ranked by the number of sources that cite it, and that count is displayed.

Each entry carries one label from a fixed list (Closest match, Cheapest, Open source, European, Easiest to switch to, Best free plan, Best for payments, Most integrations, Best for surveys). A label states why a tool is on the list and has to be provable from the recorded facts. Labels, prices and migration effort are displayed next to the similarity and never change the rank.

A sponsored entry is marked with the word Sponsored on the entry and on its link. It keeps the rank and the similarity given by the calculation: a position in the order cannot be bought. Rankings are computed before any commercial agreement.