Customer journey design: finding gaps from discovery to repeat purchase

How does journey design reveal gaps across stages?

Customer journey design starts with one person’s task and examines what they do, need to know and struggle with across time, channels and teams. It then identifies which transitions to improve. A journey is not the company’s marketing funnel or merely a website click path. People may leave a page to consult a colleague, wait for approval or compare alternatives, and those activities shape the experience. NN/g describes a journey map through an actor, scenario and expectations, phases, actions, thoughts, emotions and opportunities, retaining one primary perspective per map.[S1] This article connects that view to accountable work and measures rather than treating an emotional curve as the final deliverable.

Define actor, task and boundaries before drawing

A journey brief states the actor, trigger, goal, start and end. A new contact taking over website updates, from discovering inconsistent information to completing the first correct update, is more investigable than every customer from awareness to loyalty. Buyer, user and approver may have different journeys; do not average them into one experience. Also decide whether the map describes the present or a proposed future. A current-state map needs actual experience and records. A future-state map is a design to test. Compare them while clearly labeling the distinction so workshop aspirations do not become supposed observations. Narrow an oversized scope to a consequential task and retain other relationships as extensions.

Gather evidence of actions, waiting and experience

Reconstruct recent events through interviews, observation and available records, including online, phone, in-person and assisted interactions. Ask what happened next, how people knew the status while waiting and whom they contacted when blocked. These questions reveal mechanisms more clearly than a satisfaction score alone. Ground emotions in participants’ descriptions and context rather than adding smiley faces from team intuition. GOV.UK’s experience-mapping method begins with several actual experiences before identifying common stages, with different paths potentially requiring separate maps.[S2] One person’s account can be important without representing everyone. Record source, time, group and gaps so the map shows what is known instead of using visual tidiness to conceal uncertainty.

Use a task-centered journey-map structure

Use stages named after customer behavior across the map. Add rows for task, actual action, required information, touchpoint, waiting and handoff, difficulty, evidence, responsible role and opportunity. Labels such as checking suitability, preparing information and waiting for a response are more useful than internal sales codes. Attach an observation and uncertainty to each difficulty instead of placing all dissatisfaction in one cell. Define entry and completion conditions. Submitting a form does not mean a request has been handled, and receiving an automated message does not mean a meeting is scheduled. Record returns, retries and exits. Real journeys often loop; a clean linear diagram should not erase those paths.

Distinguish an interface issue from an operational handoff

The same difficulty can have different causes. Unclear progress may come from missing page information or from nobody owning the request internally. Trace where status information is created, who confirms it, when it can be shared and how exceptions work. If the cause is backstage, connect the journey to a service blueprint or process improvement instead of adding a more attractive progress bar. Prioritize customer-task impact, observed coverage, confidence and dependencies. A small copy fix may be a useful test but cannot substitute for an unestablished promise. Cross-team work needs a shared completion condition, such as a responsible contact receiving the request and being able to explain the next step.

Illustrative scenario: the gap after submitting an enquiry

Imagine an easy-to-use enquiry form that only says success after submission. The reader does not know whether anyone received it, what to prepare or when to check. Research might show repeated emails while sales treats them as new enquiries. Improvement requires more than better confirmation copy: define received, assigned, needs information and responded states with ownership. Rehearse a small case involving submission, assignment and another colleague taking over. Check whether information travels and duplicate requests can be recognized. Any response-time promise must match operational capacity. Do not invent a guaranteed deadline to complete the diagram. This is a hypothesis for investigation, not a claim that the current website has implemented such backend behavior.

Pilot the change before redesigning the whole journey

For each improvement hypothesis, record the affected stage, change, group, collaborators, test method and stop condition. Start with one consequential break, such as explaining the next step after an enquiry, before expanding. A clearly labeled rehearsal may be enough to test a process; it does not need a production system first. Passing that rehearsal is not deployment. Keep comparisons interpretable through similar tasks, recruitment conditions and questions. Check whether the change merely moves a problem downstream, for example a shorter form creating more follow-up requests. Consider total effort for customers and staff as well as completion. Shifting work to customers should not be reported as better experience simply because internal handling time falls.

Which measures belong in journey acceptance?

Choose task-specific measures: end-to-end completion, waiting between stages, repeated explanations or submissions, comprehension of critical information and reasons for seeking help. Define starts, ends and denominators. Completion might mean eligible people who began a task and finished it, not completed tasks divided by page views. Classify exits: recognizing that an offer is unsuitable can be a correct outcome, not necessarily design failure. Deliver scope, current and candidate future maps, an evidence index, breaks, ownership, hypotheses, pilot results and untested areas. Follow an actual task from start to finish and have relevant teams confirm the handoffs. A display map without responsibility for the next action has not yet become improvement.

Frequently asked questions: funnels, multiple actors and emotions

Can a journey map replace a marketing funnel? They answer different questions. A funnel examines selected transitions; a journey examines a person’s task and cross-channel experience. Connect them without replacing customer actions with company stages. Should a multi-person decision use one map? Preserve each role’s perspective first, then connect handoffs or relationships rather than inventing an average person. Is an emotional score required? No. Actions, waiting, information and evidence can already support improvement. If emotional evidence is absent, mark it unknown rather than deciding how customers feel simply to complete the graphic.

Sources and evidence boundaries

1. S1 · Journey Mapping 101

   Sarah Gibbons, Nielsen Norman Group · Accessed 2026-10-22

   Supports the cited principle; operational steps and illustrative scenarios are editorial recommendations rather than reported customer results.

2. S2 · Creating an experience map

   UK Government Digital Service · Accessed 2026-10-22

   Supports the cited principle; operational steps and illustrative scenarios are editorial recommendations rather than reported customer results.