How we research

We use the free account

For most profiles we sign up for a free account and spend about an hour on the tool with a test resume that has personal details removed. We try the main tasks, note what's free and where the paywall starts, and save screenshots. We stop at the upgrade or checkout screen.

We haven't paid for any plan so far. Paid features are described from the vendor's own pages and from what users report, and the profile says they weren't tested. When we haven't used a product at all, its first screen says so.

Prices: what you're charged, not the headline

We record the amount you're charged and how often, in US dollars, with the date we saw it, on the pricing page and in the app when prices appear only after sign-up. A per-month figure is there only to compare plans.

We also note what a pricing page doesn't make obvious: weekly plans that cost more over a month than the monthly plan, a plan selected by default, discount pop-ups, and products sold as separate subscriptions.

Terms: trial, refunds, renewal and cancelling

We read the vendor's terms and refund policy and write the conditions as possibilities, not promises: "you may be eligible for a partial refund", never "you'll get your money back". When the vendor's own pages disagree, we follow the binding document, usually the terms, and say that the pages disagree.

We haven't asked for a refund or cancelled a paid plan ourselves. Where that matters, the profile says it isn't checked.

What users say

We read reviews from the past 12 months, mainly on Trustpilot, plus Reddit threads where people discuss the tool, and group what comes up again and again. Praise and complaints follow the same rules: the same period, the same exclusions and at least three mentions before something counts as a theme.

• We leave out ads, posts by competitors and accounts linked to the vendor.
• Reviews the company invited are labelled, not removed. Invited doesn't mean fake.
• A Trustpilot rating covers the whole company, so we show it with the number of reviews; a rating from a handful of reviews says little.
• App and browser extension store ratings are shown with their counts. We don't read store reviews by default: they mostly rate a free extension, not the product you'd pay for.

How we write it up

Every statement says where it comes from: what we saw in our account, what the vendor says (marked as the vendor's claim), or what users report. Promises such as hidden jobs or interviews within a week are attributed to the company and kept apart from what we saw.

An unknown never becomes a "no". "Not checked" means we haven't looked; "Not included" means a source says so; "Listed for Elite only" means only the vendor's plan list says so; "Limits not confirmed" means the limit comes from a single user report.

What we don't do

• Scores and rankings. A fair ranking needs the same test on every tool, with the same resume and the same job ads. We haven't run that test, so lists are in A–Z order and say "Not ranked".
• Paid plans. Not tested so far, as each profile says.
• Results. We don't measure whether a tool gets you more interviews or a job faster.

Keeping profiles current

Prices, limits and terms change. Each profile shows when it was last updated and lists its changes under "Sources and our check"; prices and ratings carry the date we saw them. Each price is stored once and every page is rebuilt from it, so a correction reaches every page that shows it.

If you see something different from what a profile says, tell us, ideally with a link or a screenshot.

Who pays for this

No one so far: there are no ads, affiliate links or sponsored content on the site, and vendors don't see profiles before they're published. More on how Choosery is run.