CLEAR INFORMATION. INFORMED CHOICES.
Research & benchmark methodology
Standards for sources, estimates, samples and research independence.
Identify what the evidence is
Research should distinguish measured observations, self-reported survey results, modelled estimates, scenarios and editorial interpretation. A benchmark score is meaningful only with its population, method and limitations. A report labelled with a future year describes its stated outlook or coverage period; the label alone is not evidence of future observations.
Sources and reuse
For each substantive dataset or quantitative claim, record the original source, date, scope, licence or permitted use and any transformation. Link readers to primary evidence where possible. Do not treat vendor marketing numbers as independently verified findings. Respect dataset licences and do not republish restricted underlying data merely because a chart is public.
Sampling and calculations
Explain the collection window, population, sample size, inclusion criteria, missing values, units, weighting and calculation method where applicable. Do not claim representative market share or causal impact from a self-selected sample. A pasted AI answer is one observation in its provider, prompt, locale and date context; repeated generations can differ. Estimates and projections must show their assumptions and uncertainty.
Benchmark assessments
Personal assessment results are guidance based on the supplied answers and the displayed model, not a compliance certification or a guaranteed business outcome. The optional research-interest signup stores contact interest separately and does not submit assessment answers. See the Benchmark and tool methodology for the specific implementation and limits.
Independence and correction
Disclose research sponsors, their role and any restrictions. Funding must not be hidden or represented as independent peer review. Document substantive methodology changes and correct errors with their implications. This is a research standard, not a retrospective claim that every published figure has passed independent validation. Send evidence-based concerns through Corrections.
Questions, corrections or a request about your data?
Contact Markaigen