Persona Evidence Matrix¶
Documentation artifact — instantiates Evidence-Grounded Persona Proxy Design
Maps each persona claim to source evidence, assumption status, confidence, and review trigger.
A finished persona reads as a paragraph of confident-sounding sentences, and that confidence is exactly what hides how much of it is guessed. The Persona Evidence Matrix is the backstage ledger that makes each of those sentences accountable: one row per claim, with columns for the source evidence behind it, whether it is genuinely evidenced or merely assumed, how confident the team is, and what would trigger a re-check. Its single job is claim-by-claim traceability — turning "our user listens offline and won't tolerate re-buffering" from an assertion into a line item you can follow back to the interview it came from, the log event that supports it, or the honest label assumption. It is not the persona itself and not the persona's public scope note; it is the accountable substrate underneath the friendly face.
Example¶
A streaming-music product team has a persona, "Commuter Maya," and a slide deck full of statements about her. The matrix breaks that deck into rows. "Listens offline on the subway" → source: 5 of 12 interviews plus offline-download analytics; status: evidenced; confidence: High. "Refuses to pay for lossless audio" → source: one pricing survey; status: evidenced-but-thin; confidence: Medium; review trigger: re-check if the survey ages past a year. "Shares her account with family" → source: none; status: assumption; confidence: Low; review trigger: confirm before any family-plan work is scoped. Nothing about the persona changed, but the third claim is now visibly a bet rather than a fact — and the team stops treating "she shares her account" as a reason to build a family tier.
How it works¶
The unit is the claim, not the attribute, so a single persona field ("motivations") may spawn several accountable rows. Four columns do the work: an evidence column pointing to the specific source (or the word assumption), a status column that forces the honest evidenced/assumed distinction, a confidence column carrying uncertainty forward, and a review-trigger column that names the condition under which the claim must be revisited. The matrix is deliberately unglamorous and tabular; the payoff is that any downstream reader can audit the persona's provenance without re-interviewing anyone.
Tuning parameters¶
- Claim granularity — whether "price-sensitive" is one row or three. Finer rows expose weak evidence but multiply upkeep.
- Confidence scale — a binary evidenced/assumed flag, a High/Medium/Low band, or a numeric weight; richer scales inform better but invite false precision.
- Source-strength weighting — whether one vivid interview and 10,000 log events count the same; explicit weights curb availability-driven overweighting.
- Unsupported-claim policy — whether an assumption may stay in the persona (flagged) or must be cut; permissive keeps richness, strict keeps honesty.
When it helps, and when it misleads¶
Its strength is that it kills assumption laundering at the source: a claim cannot masquerade as evidenced when a column literally records that it is not, and the synthesis behind the persona becomes auditable rather than a black box. The classic failure is matrix theater — the ledger is filled in once for a review, then never consulted while the persona drifts on. A second trap is treating the confidence labels as measurements rather than judgments, lending numeric solidity to a soft guess. The discipline is to keep it living and wired: every review-trigger cell should actually feed the refresh mechanism, so the matrix stays a working audit trail rather than a compliance decoration. The artifact is essentially a requirements traceability matrix[n1] pointed at persona claims instead of software features.
How it implements the components¶
evidence_traceback— its core function: every important claim points to a named source or is explicitly labeled an assumption, one row at a time.synthesis_rule— the status column records, per claim, how it was derived (single interview, averaged across many, contested), documenting the synthesis rule at claim grain.uncertainty_annotation— the confidence column attaches and carries per-claim uncertainty forward into every use of the persona.
It does NOT state the persona's scope, excluded groups, or permitted uses (coverage_and_omission_boundary, decision_use_gate) — that is the job of Persona Boundary Card; the matrix is the backstage per-claim ledger, the card is the front-of-house use label.
Related¶
- Instantiates: Evidence-Grounded Persona Proxy Design — supplies the traceable evidence spine the whole proxy governance depends on.
- Consumes: Interview Cluster Synthesis — the matrix logs, claim by claim, the evidence behind the clusters synthesis produced.
- Sibling mechanisms: Interview Cluster Synthesis · Proto-Persona Assumption Workshop · Persona Boundary Card · Persona Scenario Walkthrough · Representativeness Review Checklist · Counterpersona Review · Persona Refresh Trigger
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Persona Evidence Matrix operates as a static representation, map, specification, schema, or prospective plan that externalizes information because it maps each persona claim to source evidence, assumption status, confidence, and review trigger.
Independent corroboration: The frozen evidence defines Persona Evidence Matrix as 'Maps each persona claim to source evidence, assumption status, confidence, and review trigger', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Human-Computer Interaction
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Persona Evidence Matrix is rooted in human-computer interaction: UX research adapts requirements traceability to connect each persona claim to evidence and assumptions.
Related originating lineages:
- Engineering & Design — Engineering and design materially shaped Persona Evidence Matrix through reliability, physical systems, safety, and mistake-proof design.
- Ethnography & Qualitative Methods — Ethnography and qualitative methods materially shaped Persona Evidence Matrix through interviewing, situated inquiry, and cultural interpretation. Traceability from persona claims to observations and interviews supplied the evidentiary foundation.
Review resolution: Both blind reviewers agree that human-computer interaction and user-experience research is the primary origin. Reconciliation resolves alternate_origin_disagreement, domain_reach_disagreement. Formative alternate lineages are retained as engineering_design, ethnography_qualitative_methods; later breadth of use is recorded separately as domain_reach=multi_domain, while origin_mode=cross_disciplinary_synthesis describes the relationship among origin lineages.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
The matrix and the Persona Boundary Card are complements, not rivals: the matrix is the detailed provenance record a reviewer inspects, while the card is the one-glance summary a user sees at the point of decision. The review-trigger column is also the natural feed for Persona Refresh Trigger — the matrix stores the staleness conditions; the refresh rule reads them and fires.
[n1] A requirements traceability matrix links each system requirement to its origin, its design, and its tests so that no requirement is unaccounted for. The persona evidence matrix borrows the pattern wholesale, substituting persona claims for requirements and research sources for tests. ↩