Skip to content

Separate a causal question from a descriptive one

Turn an ambiguous request about impact into a clear distinction between observed discussion and an effect the available evidence cannot establish.

Guides

Can public discussion show that a launch caused a change?

A request to explain whether a tutorial launch reduced setup problems contains a causal claim. Public posts may reveal discussion before and after release, but the people posting and the visibility of their posts can change. Define what effect would need to be identified before interpreting any chronological pattern as an impact.

Example research prompt
Describe setup complaints around the tutorial release using cited public posts. Do not infer tutorial exposure or claim a causal reduction; list competing explanations.
01 / Guides

Name the missing comparison

Describe the desired contrast: setup problems with the tutorial compared with the same conditions without it. The second condition is not directly observed for the same users. Include this limitation in the research brief rather than allowing a before-and-after chart to stand in for the missing counterfactual.

02 / Guides

Recover a descriptive task

Review dated, public accounts of setup problems in a bounded corpus. Include explicit descriptions of the relevant setup step; exclude general product criticism and posts with uncertain publication dates from the chronological comparison. Keep unknown exposure to the tutorial as unknown, even when a post appears after release.

03 / Guides

Inspect competing changes

Look for documented software updates, onboarding redesigns or changes in the selected accounts during the same period. Record them as plausible competing explanations, not measured adjustments. A reader mentioning the tutorial does not prove that other readers saw it or that the tutorial caused their reported experience.

04 / Guides

Use an honest conclusion

Deliver a descriptive timeline and a separate list of evidence needed for an impact study, such as an authorized experiment or a suitable comparison design. If the owner still requires a causal answer, label it unresolved. Do not convert repeated anecdotes into an estimated effect simply to complete the requested narrative.

Keep the boundaries clear.

  • Chronology can motivate a causal study without supplying its missing comparison.
  • Platform badges indicate research context, not guaranteed source access. Unavailable evidence remains unavailable.

Start with a read-only check: ask your agent to report actual source coverage before collecting records or exporting an audience.

Primary references

Documentation reviewed 2026-11-13. Product capabilities, client availability and plans can change. Third-party marks identify their products, not an affiliation.

Know what social media shows about you.
Start with a question.

Get started
montage