Research methodology

Study design

Each study states its question before data collection begins. We publish the sample definition, collection window, and unit of analysis alongside the results so readers can judge fit to their own situation.

Sampling

Samples are described by specialty mix, organization size, and geography. Where a sample is not representative of the wider market — and it usually is not — we say which direction the bias likely runs.

Normalization

Financial figures are normalized for currency and reporting period. Performance metrics are defined explicitly, because terms like “lead”, “conversion”, and “new patient” are not used consistently across the industry.

Statistical treatment

We report medians and interquartile ranges rather than averages for skewed distributions, and we do not report differences as meaningful without stating the sample size behind them. Small-cell results are suppressed rather than published as precise-looking noise.

Limitations

Self-reported data carries recall and incentive bias. Platform-sourced data reflects platform definitions. Every study carries an explicit limitations section; treat findings as directional benchmarks, not targets.

Citation and reuse

You may cite our findings with attribution and a link to the original study page. Please do not reproduce full charts or datasets without permission, and do not present our figures as your own primary research.