Select an editorial fork before collecting attributes
Choose one decision that might change by audience situation: opening explanation, evaluation criteria, evidence depth, format, call to action, or post-purchase support. Write the cost of maintaining separate paths and the condition for merging them. Do not begin by listing every demographic field available in analytics. The workflow stops with a small set of cards that produce distinct, testable editorial choices; it does not continue until the audience has been divided into a visually complete persona deck.
Evidence: Government Digital Service; IT University of Copenhagen Research Portal
Recruit around recent situations and outcomes
Find actual or likely users who recently attempted the relevant progress, including those who delayed, abandoned, improvised, or chose an indirect alternative. Ask about the triggering event, sequence, trade-offs, anxieties, and what counted as a satisfactory outcome. The Christensen perspective emphasizes circumstances behind behavior change, while GOV.UK requires user needs to rest on research. Protect consent and avoid inferring sensitive traits. Record selection boundaries so the team knows who the cards do not represent.
Evidence: Clayton Christensen Institute; Government Digital Service
Extract jobs without turning solutions into needs
Write the desired progress in the participant's context, not as a demand for the team's product. 'I need an email sequence' is a solution; 'I need to remember a time-sensitive renewal without checking a portal every day' is closer to a job. Separate functional progress, emotional reassurance, and social consequence only when evidence supports them. Note current alternatives and why they were chosen. HBR's Jobs-to-Be-Done account helps expose competition across product categories, but every local job statement still requires local evidence.
Evidence: Harvard Business Review; Clayton Christensen Institute
Cluster by causal similarity and preserve edge cases
Group episodes when trigger, desired progress, barrier, and success criterion imply the same editorial response. Do not cluster solely because people share age, occupation, or a quoted phrase. Mark cases that fit two groups and look for a circumstance that separates them. The segmentation research critique supports making the grouping logic explicit rather than treating segments as discovered objects. Give each cluster a confidence level and cite the episode IDs; a memorable name is optional and must not introduce stereotypes.
Complete a decision-linked card
Use fields for situation, trigger, job statement, current alternatives, barriers, anxieties, evidence need, successful outcome, observed episodes, disconfirming case, excluded context, and date. Then add the essential field: editorial consequence. Specify which claim, example, comparison criterion, tool, route, or next action changes. If the same article would serve two cards without meaningful adaptation, merge them. If adaptation creates contradictory advice, design a clear reader self-selection cue instead of silently assigning visitors.
Evidence: Government Digital Service; Clayton Christensen Institute
Test one variation without profiling individuals
Create a small content change that lets readers choose by situation—for example, separate apartment-noise and heavy-training decision paths—rather than predicting identity from hidden attributes. Define usefulness evidence such as successful navigation, resolved questions, or reduced confusion. Do not claim causal lift from a tiny uncontrolled sample. Review accessibility, privacy, and whether the branch disadvantages people who do not match the assumed pattern. Preserve the original version and contrary feedback.
Evidence: Government Digital Service; IT University of Copenhagen Research Portal
Retire cards that stop changing the work
Review after new research, product changes, channel shifts, or repeated ambiguous assignment. Merge cards whose consequences converge, split only when a recurring circumstance produces a new decision, and archive unsupported groups with reasons. Jobs and alternatives can change even when demographics remain stable. The deck is a working decision system, not a portrait gallery or a claim that every reader follows one job. Assign an owner and next review before the team builds more content around the segmentation.
Evidence: Clayton Christensen Institute; Harvard Business Review
Run an assignment walkthrough before publishing branches
Give several anonymized episodes to an editor who did not build the deck. Ask which card applies, which evidence supports that assignment, what happens when two fit, and what the inclusive default provides. High disagreement signals unclear boundaries or unnecessary splits. Revise the cards before personalizing a live experience. The walkthrough tests whether the model is usable and reviewable; it does not prove the groups exist at population scale.
Evidence: IT University of Copenhagen Research Portal; Government Digital Service
Sources and further reading
These references informed this article. A source supports a claim; it does not imply endorsement of TenMultigure or any future product reference.
- Our Theories: Jobs To Be DoneClayton Christensen Institute · Accessed August 10, 2026
Provides the behavior-change and circumstance lens used to recruit episodes and write solution-independent job statements.
- Learning about users and their needsGovernment Digital Service · Accessed August 10, 2026
Supports evidence-based needs, full-context research, and explicit links from segment cards to practical content choices.
- Know Your Customers' Jobs to Be DoneHarvard Business Review · Accessed August 10, 2026
Supplies the primary Jobs-to-Be-Done explanation used to identify indirect alternatives and non-product competition.
- Is segmentation a theory? Improving the theoretical basis of a foundational concept in business-to-business marketingIT University of Copenhagen Research Portal · Accessed August 10, 2026
Provides independent grounds for documenting clustering logic, uncertainty, and the limits of segmentation as theory.
Reviewed by TenMultigure Editorial Review. See an error or a source that has changed? Tell the editorial team.
Review method: AI-assisted desk research with editorial checks. Reviewed ; next scheduled review . Turned JTBD segmentation into a seven-step card-building workflow, with recent-situation recruitment, causal clustering, content consequences, self-selection, and retirement tests.