How this research gets made

SaaS Research Lab reads SaaS products the slow way. Pricing pages, signup flows and calls to action are opened by hand, scored against the same 24-field rubric, and re-checked when a product changes. Nothing here is scraped once and left to rot.

The sample

The dataset covers 1,000+ SaaS products across 38 categories. Products enter the sample when they have a public pricing page and a self-serve or demo path. Each record carries the acquisition model, tier count, starter price, annual discount, trial length and the exact wording of the primary call to action.

The scoring

Every field has a written definition, and two readers score anything ambiguous. Where a product is genuinely unclear — a hidden enterprise tier, a trial that behaves like freemium — it is marked as such rather than forced into a bucket. Reported figures always name their sample size.

The experiment network

47 SaaS sites run shared A/B tests through the network. One snippet, one test at a time, results published back to every participant. Tests run for a fixed 14 days so nobody stops a test the moment it looks good.

Citing or challenging the work

Cite anything here with a link. If a figure looks wrong, say so — write to hello@saasresearchlab.com with the product and the field, and corrections get published with the next update.