How we compare software

What we check, where the numbers come from, how the ranking is set, and the things we will not do to fill it.

comparisons: 4comparisons
tools compared: 20tools compared
individual reviews: 20individual reviews
sponsored placements: 0sponsored placements
tools that pay us nothing: 19/20tools that pay us nothing

Who writes this

ToolSieve is run by a small editorial team that does not take payment for placement, does not accept free accounts from vendors, and publishes what each vendor pays us. We do not put our names on the site because the prices should be checkable without trusting anyone. Every figure links to where we checked it.

That is a deliberate trade. Every large comparison site in this category runs on a named reviewer with a photograph, and that byline is the trust mechanism. We have replaced it with two things a reader can verify without us: a dated record of every change, and a ledger naming what each vendor pays.

The four rules

1.Prices come from the source. Every price is read on the vendor's own live pricing page or official documentation, in the currency that vendor quotes, and each comparison prints the month it was last read. Where a vendor publishes only adjacent tiers, we interpolate and mark the figure with a tilde rather than presenting it as read.
2.The renewal price is a headline criterion. An intro price is a customer-acquisition cost; the renewal is the real price. Where a vendor publishes one, it gets its own column and the multiplier is printed beside it. Where a vendor publishes none, the column says so instead of carrying an estimate.
3.Every tool gets a reason not to buy it. Including the ones we rank first, and including the one in the verdict box before you scroll. A recommendation without a boundary is an advertisement. Weaknesses are compiled from user reviews, vendor terms and community complaints, never from a vendor's own comparison chart.
4.The incentives are published, tool by tool. Each comparison carries a ledger naming what every vendor on it pays us, including the ones that pay nothing, and the rank each one holds. The rank order is set before that column is filled in.

What we commit to, specifically

These are claims, not adjectives. Each one is either true or it is a bug you can report.

We have never been paid for placement, in any comparison, in any category.
We did not receive complimentary accounts from any vendor on this site. Where a free tier was used, it is the same free tier a reader gets.
No vendor sees a comparison before it is published, and no vendor can edit one after.
We list at least one tool per guide that pays us nothing, and we mark it beside the link rather than in a footnote.
We publish tools with no affiliate programme at all when they belong in the comparison. Mullvad and Attio can never pay us anything and both are on this site.
We do not run advertising, we do not sell placements in the newsletter, and we do not publish sponsored reviews presented as editorial.
When a figure turns out to be wrong, we change it and record the change on the changelog with a date.

What the score means

Every rating on this site is an editorial score out of ten, computed by nobody: it is our summary of the criteria each comparison publishes, in the order that comparison lists them. It is not an average of user reviews, because we have no users to average and would not invent them. The bands are printed beside every score on the site, and they are these.

8.0 and up: Strong. Wins its category on the criteria below, with a known trade-off.
7.0 and up: Solid. Right answer for a specific reader, wrong one for others.
6.0 and up: Adequate. Works, and something on this page does the same job better.
under 6.0: Weak. We would not buy it at this price.

Corrections

Vendors change prices without telling anyone. If a number here is stale, tell us and we will fix it and record the change. Everything we have changed so far is on the changelog, and the money side is on the disclosure page.