How the guides are researched

Features, prices and plan limits come from the company that makes the tool: its help docs and its pricing page. Nobody documents a product in more detail, and nobody has a stronger reason to describe it kindly.

So the vendor's word is used for what a product does and costs, and not much else. When a page says one tool suits you better than another, or leans on a benchmark, that part needs support from someone who isn't selling anything.

Every article ends with a list of the sources it was written from, with the date each one carries where it gives one. Links in the text go to the page behind that sentence. Each article shows when it was published and when it was last updated.

Before a page goes up, every source behind it is opened again and compared with what the page says. A wrong figure, a missing condition, or a "most" where the source only said "many" gets fixed first.

The guide shows both sides instead of picking one. The average order value guide is a good example. Four publishers say cancelled, unpaid and pending orders shouldn't count. The store platform's own reports count them.

They count different orders, so two AOV figures can't be compared until you know which rule each one follows. The guide lists what to write down beside the number before you compare it.

Without scores. A 7 out of 10 folds a dozen separate judgements into one number, and two tools with the same score can be wrong for you for different reasons: one charges by contact count, the other's automation builder is too basic for your flows.

Instead, each tool gets who it is built for, where it is strong and where it falls short, and the weak points need the same sourcing as the strong ones. Lists are organised around why you might be looking, because a store leaving a platform over price needs a different answer from one that needs SMS.

Prices and plan limits go out of date faster than anything else, so tool pages state the day their prices were read. A tool page is due for a full refresh six weeks after its last update: every source read again, prices and plans corrected, and the list itself reconsidered. The updated date only moves when that has actually happened.

A figure that can't be traced back to a source. A benchmark with no known sample, date or method. A topic that credible sources can't answer properly gets narrowed to the part they can, or it doesn't get written.