How we test and review software

The short version

We make money when you buy software through our links. That is a real conflict of interest, and no amount of "we value integrity" boilerplate makes it disappear. What follows is the set of rules we work under to keep that conflict from bending our conclusions — and, just as importantly, the things we don't do, so you're never misled about how much testing sits behind a recommendation.

1. How we choose what to cover

We cover categories where the buying decision is genuinely hard and the wrong choice is expensive to undo — email platforms, security tools, course and community platforms, the software that a business builds a workflow on top of. We start from the question a real buyer types into Google, not from whichever program pays the highest commission.

A tool makes it into a roundup because it is a credible answer for some reader, not because it has an affiliate program. When a category's best option has no affiliate program at all, we still name it — and we say plainly that we earn nothing from it. You will see this in our roundups; if you never do, that is a reason to stop trusting us.

2. What "hands-on" means here — and what it doesn't

We label every tool in every review with the depth of our own experience with it, using three levels. We never blur them:

• Used in production. We have run real work through this tool — real subscribers, real campaigns, real customer data — for a meaningful period. We can tell you how it behaves when something goes wrong.
• Hands-on tested. We created an account, built the core workflow the tool is sold for, and pushed on the specific things buyers ask about. We have not lived with it for months.
• Researched, not tested. We have not used it ourselves. Our assessment comes from official documentation, current pricing pages, feature limits, migration and cancellation terms, and a structured read of verified user reviews on G2, Capterra and TrustRadius — weighted toward recent reviews and toward complaints that repeat.

The third level is still useful work — pricing traps and plan limits are documented facts, and you can check them yourself from the sources we cite. But it is not the same as having used the product, and we refuse to write it up as though it were. A tool at level three never wins a roundup on the strength of our opinion alone; it wins on verifiable specifics, or it doesn't win.

3. How we verify pricing

Every price, plan limit and contact tier in our tables is read off the vendor's own pricing page and stamped with the date we checked it. We do not quote prices from memory, from other blogs, or from the vendor's affiliate materials — those are the three places stale pricing comes from.

Where it matters to the decision, we also record the things vendors put a click or two away from the pricing page: what the price becomes at renewal, whether the headline figure requires annual prepayment, what happens to the bill as your list or seat count grows, and whether you can export your data and leave without a support ticket. SaaS pricing changes often; if you find a figure of ours that has gone out of date, tell us and we will correct it and re-date the page.

4. What we actually compare

Feature checklists are the laziest form of software review — every tool ticks every box, and the boxes don't predict whether you'll be happy in a year. We weight our assessments toward the things that actually decide it:

• Cost at your real size, not at the marketing tier — including the step where the price jumps.
• Deliverability, reliability and support — does the thing work when it matters, and can you reach a human.
• The exit. How hard is it to export your data and migrate away. A tool that traps you is a worse buy at any price.
• Who it's genuinely for. Most category winners are wrong for a large share of readers. We say which share.

5. How commissions are kept out of the verdict

Three rules, and they are checkable against the published pages:

• The ranking is written before the affiliate links are added — never adjusted afterward to favour a better-paying program.
• We do not accept payment for placement, ranking, or a positive review. No vendor gets to see a review before it publishes, and no vendor gets approval over what we write.
• We publish real drawbacks for every tool, including our top picks. A review with no "cons" section is an advertisement, whatever it calls itself.

We also name the commission structure when it's relevant to judging us. If we rank a tool first and it happens to pay us a lifetime recurring commission, you deserve to know that while you read — not buried in a footer.

6. When we get it wrong

We will. Pricing changes, products get worse, a tool we recommended gets acquired and gutted. When that happens we update the page and note what changed and when, rather than quietly editing the text. Reviews here carry a "last updated" date for that reason — an undated software review is close to worthless.

If you bought something on our recommendation and it didn't work out, we want to hear it. That feedback is the only real quality control a publication like this has.

---