How we test, score and pick

How a review is made

1. We choose what to review by what small businesses actually buy. A tool earns a review by being one a reader is likely to be choosing between, not by asking for one. Vendors cannot buy a review, and a vendor asking for one does not move it up the queue.
2. A staff member signs up for a fresh account. The free plan or the trial, exactly as a reader would get it. Where a vendor offers a demo account instead, we say so on the page, because a demo is not the product a reader will be paying for.
3. We run the tool against the five criteria below. The same five on every review, in every category, so two reviews can be read side by side. Notes are taken as we go: what worked, what was confusing, what only appeared after an upgrade prompt.
4. We ask support a real question. One that a new small-business user would ask, on the channel the plan we are on actually gets. We record how long the answer took and whether it was right.
5. We read the pricing page, not the press release. Every plan line on the site is checked against the vendor's own current pricing page. We describe the shape of the pricing rather than quoting figures, because the figures change and a stale number is worse than none.
6. A second member of staff reads it before it goes up. They check every factual claim against the product and every score against the notes. Where they disagree with the reviewer, the disagreement goes in the review.
7. We keep it current. Pricing pages and product updates are re-checked, and the review's updated date changes only when something in it did.

The five criteria

Each is scored from one to five, in whole and half points, by the reviewer. The overall score is the mean of the five, to one decimal, with equal weight: we do not tune the weights per category, because the point of a shared scale is that it is shared.

• Ease of use. How quickly a new user is doing real work, how much has to be learned first, and whether the everyday screens (the pipeline, the editor, the board) stay out of the way once learned.
• Features. Whether the core job of the category is covered on the plan a small business would actually buy. A feature that only exists one tier up counts against, not for.
• Value for money. What the paid tiers buy relative to the free tier or trial, how the bill grows with seats or contacts, and how it compares with the other tools on the same shortlist. We judge the shape of the pricing, not a figure that will be out of date next quarter.
• Support and docs. The help centre, the channels a small plan gets, how long a real question took to answer during testing, and whether the answer was right.
• Fit for small business. How well the tool fits a solo operator or a team under twenty, which is who this site is written for. An excellent enterprise product can score low here and that is the point.

What the scores mean

• 4.5 and above. We would choose it ourselves for the job it is on the shortlist for, and the reasons to skip it are about fit, not quality.
• 4.0 to 4.4. A good tool with a real caveat. Most of the shortlist lives here, and the review's "skip it if" line is the part to read.
• 3.5 to 3.9. Right for a narrow set of readers. It is on the site because that set exists, not because it is a safe default.
• Below 3.5. We would not put it on a shortlist. It may still have a review, so that the reasons are written down somewhere.

How shortlists are made

A best-of list starts from the reviews. We take every tool we have reviewed in the category, order them by score, and then decide the list by fit: a shortlist for a small business does not need the highest-scoring enterprise tool on it. Lists stay short, two to five picks, and a tool comes off a list when it stops earning the slot, not when a better affiliate rate appears.

An alternatives list starts from one tool and the reasons readers give for leaving it, then lists the reviewed tools that answer those reasons. A head-to-head compares two tools on the dimensions that actually change the decision, and ends with a straight answer about which to pick.

What we do not do

• We do not take payment for a review, a position or a score. Vendors cannot sponsor a shortlist, buy a higher slot, or pay to change a word of an entry.
• We do not publish reader ratings. Every score on the site is a staff score with a name on it. We may add reader feedback one day; when we do, it will be labelled as such and kept separate from the editor's score.
• We do not invent figures. No testing-hours counter, no "trusted by" number, no price we have not read on the vendor's page that week. Every count on the site is computed from what is actually published.
• We do not use vendor-supplied copy or screenshots. What a page says about a tool is what the reviewer saw in their own account.

How we make money

Affiliate commissions. Some of the vendor links on this site carry an affiliate tag, and if you sign up through one, the vendor pays us a commission at no extra cost to you. That is the whole business model, and it is stated on every page that has a vendor link, at body size, directly under the title.

Commissions differ between vendors, and we know what they are. The way we keep that from moving a score is structural: the reviewer scores the tool before anyone looks at the programme, the second reader checks the score against the notes, and a tool with no affiliate programme at all is reviewed and listed the same way. Theaffiliate disclosure has the full statement.

Corrections

If a review is wrong about a fact, a pricing line has gone stale, or a "skip it if" is unfair to your experience, tell the desk. Factual corrections are made on the page, with the updated date changed, and a correction that changes a score changes the score.