Journey Mapping And Service Design

Journey Mapping & Service Design

Use When

• You need an evidence-based persona — a research-grounded archetype with goals, context, and jobs — not a demographic stock photo invented to justify a roadmap.
• You must map a user's or customer's experience over time across stages, touchpoints, and channels: a current-state journey map, a future-state map, or a broader experience map (organisation-agnostic, before a specific product exists).
• You are designing or fixing a service, not just a UI — you need a service blueprint that ties what the customer sees (frontstage) to the staff actions, systems, and processes that deliver it (backstage + support), separated by the line of visibility.
• You need to express what the customer is fundamentally trying to accomplish as Jobs-To-Be-Done ("when I ___, I want to ___, so I can ___") so the design serves the job, not the feature.
• You have research and a map and must produce an alignment diagram (research → map → opportunities) and a ranked opportunity backlog that the team can act on.

Do Not Use When

• You are running the interviews, surveys, or usability tests that feed the map → use the sibling ux-research-and-usability-testing (that skill gathers the evidence; this skill structures it over time). The two pair — see Workflow steps 1 and 8.
• You are designing a single screen's structure, fidelity, or flow → use the sibling wireframing-and-prototyping (a wireflow is one touchpoint's mechanics; a journey map is the experience around it).
• You are applying cognitive/behavioural theory or heuristics to a design (biases, cognitive load, Gestalt, Nielsen) → use ux-psychology.
• You are defining the end-to-end product/discovery operating model, RACI, and rituals → use enterprise-ux-process.

Required Inputs

• The decision or problem the map must inform. A map with no decision behind it is a poster. Name it: which journey, for whom, to resolve what (drop-off? a broken hand-off? a new service we are scoping?).
• Evidence. Research notes, interviews, support tickets, analytics funnels, frontline-staff input, complaint logs. Personas and emotion curves invented without evidence are slop (see Anti-Patterns).
• The actor and the scope. Who the journey belongs to (one persona, named), and the start/end boundary — where the journey begins (often before your product) and where it ends (often after).
• For a blueprint: the people, systems, channels, and back-office processes that deliver the service — including the parts the customer never sees.
• The channels and touchpoints in play (app, web, phone, in-person, SMS, third parties), because cross-channel seams are where experiences break.

Workflow

1. Anchor to a decision and a persona. State the decision the map will inform and whose journey it is. Build the persona from evidence per references/journey-map-templates.md §Persona — goals, context, jobs, real quotes, and the research behind each. If you cannot cite evidence for a persona trait, mark it an assumption to validate (route to ux-research-and-usability-testing). No invented personas.
2. Frame the Jobs-To-Be-Done. Before mapping the how, state the why as JTBD: "When [situation], I want to [motivation], so I can [outcome]." The job is stable; the solution is not. Use references/journey-map-templates.md §JTBD. The journey then serves the job, so opportunities are judged by "does this advance the job", not "is this a nice feature".
3. Choose the map type. Journey map (one persona, one scenario, your service exists), experience map (broader, organisation-agnostic, problem space before a solution), or service blueprint (you must change how the service is delivered, not just the interface). references/journey-map-templates.md §Choosing the map names which to use when. Picking the wrong artifact is the most common waste.
4. Lay the spine: stages and actions. Break the journey into the actor's stages (their mental phases, named in their words — not your funnel), then the actions/touchpoints within each. Use representative content and real touchpoint names; never lorem, never generic "Awareness → Consideration → Decision" if the real journey isn't that.
5. Add the experience layers. For each stage capture: what they're doing, thinking, feeling (the emotion curve, justified by a quote or datapoint — not a decorative squiggle), the touchpoint/channel, and the pain points & moments of truth. An emotion curve with no evidence under each peak/trough is decoration; cite the source.
6. For a service, blueprint it. Add the swimlanes below the line of interaction: frontstage staff actions, the line of visibility, backstage staff actions, the line of internal interaction, and support processes/systems. Use references/service-blueprint.md. The value of a blueprint is precisely the backstage failure points the customer never sees but always feels — find them.
7. Mark the seams and moments of truth. Cross-channel hand-offs, waits, and ownership gaps (where no team owns the step) are where services fail. Flag each on the map. A blueprint that hides the hand-offs has skipped its only real job.
8. Build the alignment diagram and rank opportunities. Turn every pain point / unmet job into an opportunity, traced back to its evidence (research → map → opportunity — the alignment-diagram spine). Rank by impact on the job × frequency × feasibility. Use references/journey-map-templates.md §Opportunity backlog. An opportunity with no evidence behind it, or a pain point with no opportunity in front of it, is incomplete.
9. Apply the doctrine lens. A journey where the customer says "it felt generic / I didn't trust it / it looked like everyone else's" is a slop signal in experience form (doctrine/design-doctrine.md §0, doctrine/references/ai-slop-taxonomy.md) — the moat is the experience reading as deliberately authored. Escalate those troughs as design opportunities, not nuisance comments.
10. Check accessibility and channel-equity coverage. If the map assumes one happy channel and never models the keyboard-only, screen-reader, low-bandwidth, low-literacy, or offline-fallback path, state that as a known gap and pair with doctrine/references/wcag-2.2-criteria.md before claiming the journey "works for everyone".

Decision Rules

Capability Contract

• Must read traceable research and operational evidence; synthesis is read-only unless workshop or artefact editing is requested.
• May facilitate mapping, but may not invent user evidence, assign organisational commitments, or publish sensitive maps without authority.

Degraded Mode

• If actor, scope, or evidence provenance is missing, stop the final map and return a proto-map with explicit assumptions.
• Without workshop access, produce an asynchronous draft for owner validation. Recover conflicts by preserving competing evidence, naming the owner, and marking unresolved branches.

Quality Standards

• Every major stage, pain point, and opportunity traces to evidence or is labelled assumption.
• Opportunities identify affected actor, causal layer, owner, expected outcome, and validation need.

Anti-Patterns

• The invented persona. A stock-photo archetype with made-up demographics and a cute name, built to rationalise a decision already made. Personas not traceable to research are confirmation laundering.
• The decorative emotion curve. A smooth sad-to-happy squiggle with nothing under each peak/trough. If a low point has no quote, ticket, or metric beneath it, it is fiction.
• Happy-path-only journeys. Mapping only the smooth path and omitting errors, waits, recovery, and the channel switch — hiding exactly the moments where the experience breaks.
• The wall-poster. A beautiful laminated map with no decision, no owner, and no opportunity backlog. It gets framed and never changes anything (the experience-scale cousin of the shelved research deck).
• Blueprint with no backstage. A "service blueprint" that only redraws the frontstage screens and never models staff actions, systems, or the line of visibility — so the failures that live backstage stay invisible.
• Stages in your words, not theirs. Forcing the customer's lived experience into the company's funnel labels, which hides where their mental model and your process diverge.
• Map as the deliverable. Treating the artifact as the goal. The map is a means; the ranked opportunities are the output. No opportunities → unfinished.
• One persona, one channel, one ability. Mapping the median user on the best device and calling the service designed.

Outputs

• One or more evidence-based personas (goals, context, JTBD, real quotes, traceable to research).
• A set of JTBD statements framing the journey by outcome, not feature.
• A journey map / experience map: stages (in the actor's words) × layers (doing / thinking / feeling / touchpoint / pain & moments of truth), with an evidence-justified emotion curve.
• Where a service is involved, a service blueprint: frontstage / line of visibility / backstage / support processes, with seams and failure points marked.
• An alignment diagram + ranked opportunity backlog: each opportunity traced to evidence and scored by impact × frequency × feasibility.

Examples

• See examples/journey-map-and-opportunities.md — a worked, evidence-based current-state journey map for Maduuka (a B2B inventory app for small East-African retailers): the persona, the JTBD, the stage-by-stage map with an evidence-backed emotion curve, the cross-channel seams, and the ranked opportunity backlog traced back to research. Representative content throughout — never lorem.

References

• references/journey-map-templates.md — the persona template, the JTBD format, the journey-vs-experience-map decision, the stage × layer map grid, the emotion-curve evidence rule, and the alignment-diagram / opportunity-backlog scoring model.
• references/service-blueprint.md — the blueprint swimlanes (frontstage, line of visibility, backstage, line of internal interaction, support processes), how to find backstage failure points, and how a blueprint differs from and extends a journey map.
• doctrine/design-doctrine.md — §0 Mission (the authored-not-templated moat) and §2 Anti-Slop Charter; an experience that "feels generic / untrustworthy" is a slop signal at journey scale.
• doctrine/references/ai-slop-taxonomy.md — the slop tells; map "felt generic / didn't trust it" troughs here and escalate them as opportunities.
• doctrine/references/wcag-2.2-criteria.md — accessibility/channel-equity gate; a single-channel, single-ability map is incomplete, not done.
• Sibling skills: 05-ux-process-research-and-psychology/ux-research-and-usability-testing (gathers the evidence this map structures), wireframing-and-prototyping (the touchpoint mechanics within the journey), ux-psychology (theory behind the emotion curve), enterprise-ux-process (the surrounding operating model).