Start with the question
Every piece of work should begin with a defined question, an explicit scope and a reason it matters. A broad industry claim cannot rest on evidence that only describes one platform, market or group.
Make the source visible
Research entries should identify the original publisher, publication date, geography and relevant definitions. Primary sources should be distinguished from commentary and secondary reporting.
A reader should be able to trace a claim to its source and understand whether the underlying evidence is available for inspection.
Explain the method
Original research should describe how observations were collected, how the sample was selected and how the analysis was carried out. Where practical and permitted, supporting materials should be made available.
Third-party research should be accompanied by a plain-language account of its method, as far as the publisher discloses it.
Keep uncertainty in view
Estimates, reported figures and projections should be labelled distinctly. Limitations, missing information and differences in definitions should be visible alongside the findings.
Comparisons should only be made where the underlying measures support them. The absence of reliable evidence is itself worth acknowledging.
Review and correct
Published work should have an identifiable review process. Material corrections should be dated, explained and linked to the affected publication.
Methods will evolve as the observatory develops. Changes that affect interpretation should remain visible to readers.
