Overlap and Coverage Dashboard¶
Analytics dashboard — instantiates Inclusive Membership Union Design
Renders the union's overlap structure and per-source coverage at a glance, and stamps the standing warning that any-source membership is not validation.
A materialized union hides its own shape: the single flat list does not say how much of it came from each source, how much the sources overlap, or how much any one source adds that no other did. Overlap and Coverage Dashboard is the read-only view that makes that structure visible. It profiles the current union — how many members each source contributed, how many are shared versus unique, what fraction of the whole any single source covers — and displays it so a human can see at a glance whether the union is broad-and-redundant or thin-and-fragmented. Its second job is to carry the scope caveat at the point of consumption: right where someone reads the numbers, it stamps the reminder that a member being in the union means at least one source listed it, not that the union is a validated or agreed-upon truth set. It measures a snapshot; it does not compare snapshots over time.
Example¶
A marketing team has unioned three ad-platform audiences — a Meta lookalike list, a Google Ads customer-match list, and their own CRM segment — into one addressable audience of 480,000 people. The flat number tells them nothing useful. The dashboard breaks it open: the CRM contributes 60% of members, Google 35%, Meta 25%; the pairwise overlaps show CRM∩Google is large (the same known customers), while Meta adds 90,000 members no other source has. A coverage bar shows that dropping Meta would shrink the union by nearly a fifth, but dropping Google would barely move it because Google's members are almost all already in the CRM.
Across the top of the view sits the scope label: "Union audience — inclusion means the person appeared in ≥1 source; it does not mean they are a validated customer or opted-in lead." That one line stops a manager from reading the 480,000 as 480,000 vetted prospects. The dashboard has turned an opaque merged list into a picture of where its coverage really comes from — and a warning about what the number does and does not mean.
How it works¶
- Attribute every member to its sources. Using the folded provenance, tag each member with the set of sources that contributed it, then aggregate.
- Compute the overlap profile. Per-source counts, pairwise and higher-order intersections, and unique contributions — the full breakdown of shared versus source-exclusive membership.
- Render coverage, not just counts. Show each source's share of the union and its marginal contribution (what would be lost if it were removed), which counts alone hide.
- Pin the scope label to the view. Attach the downstream-use caveat to the display itself, so the warning travels with the numbers rather than living in a document nobody opens.
Tuning parameters¶
- Overlap depth — whether to show only pairwise intersections or full n-way overlap. Higher order reveals redundant source clusters but grows combinatorially and clutters the view.
- Coverage metric — raw share versus marginal (drop-one) contribution. Raw share is intuitive; marginal contribution is what actually answers "which source can we cut?" but is costlier to compute.
- Aggregation grain — whole-union totals versus a breakdown by segment, region, or type. Finer grain surfaces where a source dominates but multiplies the panels a reader must scan.
- Scope-label prominence — how loudly the use caveat is displayed. A subtle footer keeps the view clean but is ignored; a banner is unmissable but nags. Match it to how often the union is misread as consensus.
When it helps, and when it misleads¶
Its strength is turning an invisible structure into a decision aid: it shows which sources are redundant, which are load-bearing, and how much the union would shrink if a source dropped out — the inputs to pruning sources or justifying their cost. A natural summary statistic here is the Jaccard index, the ratio of shared to total members between two sources, which compresses overlap into a single comparable number.[n1]
Its failure mode is that a coverage picture invites over-reading. High overlap looks like corroboration — "three sources all have this person" — but the union rule guarantees nothing of the kind; overlap is just how the sources happen to intersect, not evidence of truth, which is exactly why the scope label is welded to the view. The classic misuse is trusting the profile when the underlying provenance is incomplete: if members were merged without keeping which sources contributed, every overlap number is fiction. A further trap is confusing this static snapshot with change — the dashboard says nothing about whether coverage grew or shrank since last time. The guarding discipline is to compute overlap only on members with reliable provenance, keep the scope caveat prominent, and send anyone asking "what changed?" to the version-comparison mechanism instead.
How it implements the components¶
overlap_profile— its core output: per-source counts, intersections, unique contributions, and coverage shares for the current union.downstream_union_scope_label— pinned to the view so the "membership ≠ validation" caveat is read at the moment the numbers are consumed.
This dashboard profiles a single snapshot; it does not compare versions or distinguish a real membership drop from a source outage (null_and_empty_source_policy) — that is Union Delta Review, its nearest twin: the dashboard answers "what does the union look like now?", the delta review answers "what changed since last time?".
Related¶
- Instantiates: Inclusive Membership Union Design — surfaces the archetype's overlap profile and use-scope warning.
- Consumes: Deduplicating Union Pass supplies the materialized union and folded provenance the profile is computed from.
- Sibling mechanisms: Canonical Identity Resolution Pass · Deduplicating Union Pass · Inclusive-OR Membership Test · Provenance Tagging Protocol · Type Compatibility Checklist · Union Delta Review · Union Specification Sheet
Editorial Notes¶
Form Classification¶
Form family: Interface, Display & Cue
Rationale: The mechanism is a user-facing dashboard that renders overlap, source coverage, marginal contribution, and a pinned scope warning for immediate interpretation.
Nearest alternative: Analysis, Modeling & Optimization — Metrics are computed upstream or within the tool, but the concrete deployed form is the perceptual decision surface.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Data Science & Analytics
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Overlap and Coverage Dashboard is most directly rooted in data science and analytics' computational practice of modeling, monitoring, validation, and pattern extraction. The lineage fits its defining practice: Renders the union's overlap structure and per-source coverage at a glance, and stamps the standing warning that any-source membership is not validation.
Related originating lineages:
- Mathematics — Set union, intersection, and similarity measures supply the dashboard's formal structure.
- Statistics & Experimental Design — Overlap and Coverage Dashboard also draws materially on experimental design and statistics' methods for comparison, uncertainty, sampling, sensitivity, and inferential validation, which shaped this mechanism rather than merely adopting it as an application.
Review resolution: Both independent reviews agree on primary origin data_science; reconciliation resolves alternate_origin_disagreement. Formative alternate lineages retained: statistics_experimental_design, mathematics. The broader reach of later applications is kept separate as domain_reach=multi_domain; origin_mode=cross_disciplinary_synthesis records how the formative lineages relate. Confidence is conservatively reconciled to medium, and encyclopedia_synthesis=true preserves the reviewers' boundary judgment.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; medium confidence.
Notes¶
[n1] The Jaccard index of two sets is |A ∩ B| / |A ∪ B| — the fraction of their combined membership that they share. It gives the overlap profile a single 0–1 number per source pair, but it measures co-occurrence, not agreement or validity, which is why high overlap must not be read as corroboration. ↩