What evidence would help you write content inclusion rules before collecting examples?
A guide assembled from whatever examples are easiest to find can produce a distorted picture. Decide what qualifies for a research-backed article before reviewing results. Define task relevance, source quality and exclusions, keeping unavailable coverage explicit rather than retroactively changing criteria to preserve a preferred conclusion or filling a desired number of examples with weak matches that do not actually address the question the page promises to answer.
Define inclusion and exclusion rules for this article before selecting examples, keeping hard constraints, duplicate handling and unavailable coverage explicit.Define the editorial decision
State the article's decision and the minimum facts an example must expose. Include public material with identifiable task conditions and relevant evidence. Exclude private data, unsupported claims and duplicate retellings. Record the time window and permitted source contexts, separating a hard eligibility condition from a useful optional detail so the eventual article can explain why a case was included without pretending that every missing field invalidates otherwise useful bounded evidence.
Map evidence to the reader task
Apply the rules to a small set of contrasting examples on paper. Look for ambiguity in terms such as recent, independent or relevant. Keep a reason for each exclusion and preserve unavailable material as unavailable rather than unfavorable. Do not infer missing dates, identities or outcomes from thumbnails and captions, and do not relax a user-specified constraint silently just because the current selection would otherwise contain fewer usable cases than the brief requested.
Test the angle against counterexamples
Compare the resulting set with the article's intended scope. A narrow source set may support a narrow conclusion, not a broad market statement. Seek counterexamples under the same rules rather than adding a special exception only for favorable material. If the rules yield no usable evidence, report that result and propose a clearly labeled scope change for approval instead of manufacturing examples, changing the denominator or claiming that absence proves the phenomenon does not exist.
Prepare the bounded content brief
Deliver an inclusion policy, an exclusion log shape and a bounded output specification. Keep actual collection and publication with the owner. Do not claim the rules produce a representative sample unless the method supports that assertion. The article should make selection transparent and counts exact, not imply that every relevant public account was accessible or that a curated evidence set establishes complete coverage across the platforms shown in contextual badges or mentioned in the surrounding content.
Keep the boundaries clear.
- For content inclusion rules, a proposed example is not an observed product result.
- Badges indicate context for content inclusion rules, not guaranteed access or audience coverage.
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.