How we test (and what we don't)

We have not run hands-on tests of the software we cover, and we do not publish results as if we had. Our comparisons are desk research from official pricing pages, contracts, help documentation and government sources, each with a check date.

What we do for every comparison

1. Read the vendor's own pricing page for the country we are writing about, and record the date we read it.
2. Read the contract or terms where they are public: term length, cancellation, early-termination fees, hardware ownership.
3. Read help-centre documentation for the specific feature in question (award interpretation, tip pooling, offline mode, accounting sync) and cite the article.
4. Check the rules against government sources: Fair Work Ombudsman and Fair Work Commission (AU), RBA (AU payments), US Department of Labor and IRS (US).
5. Price the scenario with a calculator that prints its formula, so you can rerun it with your own numbers.

What we leave out

• Speed, latency, sync times, tap counts and uptime. These need measurement we have not done.
• Prices the vendor only gives on a sales call. We say "quote only".
• Anything we could not trace to a source.

If we add hands-on testing

If we later run products on a real or demo account, the article will say which account, which version and on what date, and the test result will be marked separately from documentation-based claims.

See also: editorial policy, about BOHStack, affiliate disclosure. Spot an error? [email protected].