1. Home
  2. How We Test Tools

How We Test Tools

Last reviewed 12 June 2026

Every "best of" page on FreelanceToolkit UK is a recommendation, and recommendations only mean something if you understand the testing methodology behind them. This page documents how we evaluate the software, financial products and service providers we cover.

What we actually do

Our comparisons are drafted with AI assistance from providers' own published product, pricing and terms pages, and from official sources such as GOV.UK, HMRC, the FCA register and Companies House. Before a page is published, the facts it relies on (prices, plan names, regulatory status, rates and thresholds) are checked against those sources, and the page shows the date they were last reviewed.

We do not hands-on test every product, open accounts with every bank or buy policies from every insurer. Where a page describes how a product works, that description comes from the provider's published information, not from our own use. Try a free trial or ask for a quote with your own details before you commit.

Our own calculators

Calculators are checked differently: each one is run against worked examples with known answers (for example HMRC's published rates applied by hand), and the worked examples on the page come from the calculator's own output. Automated tests re-run those checks whenever the code changes.

What we score on

Fit for UK freelance use cases (specific workflows, UK regulatory compliance), pricing transparency, customer support quality (where we can assess), data portability (can you leave?), ecosystem integrations (does it play well with the rest of a UK freelance stack?), and stability (is the provider going to be around in 3 years?).

What we deliberately don't do

Assign fake "4.7/5" scores. Rank providers in a way that pretends there's an objectively-best option when the right choice depends on the reader's situation. Hide commercial relationships. Claim we have used a product when we haven't. Write "best of" lists where we're just rehashing other reviewers' opinions.

Limits of what we can test

We can't simulate enterprise-scale usage from a one-person desk. We can't run a 50-employee payroll through Xero to test how it handles. We can't test every UK lender's underwriting nuance — we rely on lenders' and brokers' published criteria for that. Where our testing is limited, we say so explicitly on the relevant page.

When something changes

Software pricing changes regularly. Affiliate provider feature sets evolve. UK regulatory bands shift each April. We refresh pages on a rolling basis — annually as a minimum, more often for fast-moving categories (banking, accounting software, outreach tooling). Every page shows a Last reviewed stamp.

Disagreement is welcome

If you've used a tool we've covered and disagree with our assessment, email us via the contact page. We read every correction request. Where readers raise material points we missed, we update the page and credit the input.

Related editorial / EEAT pages: