Skip to content

Explain overlapping category totals in a returned evidence summary

Separate category assignments from distinct evidence membership in operational tables.

Guides

Explain overlapping category totals in a returned evidence summary — what should the operator record?

A single returned object can belong to several local categories. Counting it once in each group is useful for some workflows but can make summed totals exceed distinct membership. Preserve the assignment relationship and make that arithmetic explicit, rather than editing category counts until they resemble a disjoint partition of the evidence set.

Practical operations check
Separate category assignments from distinct evidence membership in operational tables.
Use source_key, category_key and annotation_revision as a relationship table. Explain whether each pair is unique and whether category membership was source-supplied or locally assigned, keeping operational labels distinct from factual characteristics the source actually returned about the object.
Place a short measure definition beside the summary and reference the assignment revision. Recipients should understand why category sums do not equal distinct membership without treating the overlap as evidence of additional independent sources or inferring population prevalence from local tagging.
01 / Guides

Define assignment grain

Use source_key, category_key and annotation_revision as a relationship table. Explain whether each pair is unique and whether category membership was source-supplied or locally assigned, keeping operational labels distinct from factual characteristics the source actually returned about the object.

02 / Guides

Choose summary measures

Compute assignment_count and distinct_source_count separately using validated qualified keys. Neither should be labeled total people unless the actual identity grain and evidence support that interpretation; a post table or relationship expansion does not become a respondent population merely through grouping.

03 / Guides

Test overlap examples

Use a labeled fixture where one source object has two categories and another has one. Verify the category sums exceed distinct objects for a documented reason, while accidental duplicate source/category pairs are detected rather than counted as meaningful additional assignments.

04 / Guides

Preserve unassigned rows

Maintain an explicit unassigned or unresolved category state when appropriate. Do not drop eligible objects lacking a local annotation and then describe the category summary as covering all admitted evidence without stating which actual keys remain outside the assigned groups.

05 / Guides

Publish interpretable totals

Place a short measure definition beside the summary and reference the assignment revision. Recipients should understand why category sums do not equal distinct membership without treating the overlap as evidence of additional independent sources or inferring population prevalence from local tagging.

Keep the boundaries clear.

  • Separate category assignments from distinct evidence membership in operational tables.
  • Operational record: Define assignment grain. Context badges do not certify access.

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 2026-11-15. 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