In this guide
- 01Choose a work task
- 02Trace the whole experience
- 03Change the supporting work
- 04Check use and side effects
Improve employee experience by investigating a specific part of working life, changing the service behind it and checking whether people can complete their work more easily. Begin with an actual task such as obtaining equipment or arranging a shift change. A broad engagement score rarely tells you which step to repair.
This approach connects employee participation with customer-needs research, while recognizing the different power relationships inside an organization. Employees may depend on the people commissioning the research. Make participation and the use of their information clear before asking for candid accounts.
Choose a service with a visible beginning and end
An internal service usually crosses organizational boundaries. Equipment collection may involve a manager, purchasing, identity administration, IT support and facilities. The employee experiences one interrupted working day even if each department completes its own ticket on time.
The GOV.UK Service Standard recommends solving a whole problem for users, including work across team boundaries. Applied to internal services, the useful question is what the employee can accomplish at the end, rather than which department has closed its case. This is an adaptation of public-service guidance, not a requirement that every employer adopt a government process.
Choose a bounded problem. “Make onboarding better” is too broad for a first investigation. “Help a new colleague obtain a working equipment set before their first remote shift” gives the team an event, a user group and a practical outcome. It still leaves room to discover that equipment is not the main barrier.
Trace one internal service from beginning to end
Consider a hypothetical organization in which new employees collect laptops from a support counter. The collection ticket is marked complete, but a colleague discovers at home that the charger does not fit. Their manager records a late start; IT records an equipment issue. Neither record shows the whole experience.

Trace the sequence from the equipment request to the first completed work task. Ask employees to describe a recent occasion, including what they were trying to do, whom they contacted and where they waited. Include service staff: they may be working around the same fragmented systems.
Contextual research guidance describes observing people in their everyday environment. For internal services, that could mean watching equipment collection with permission or examining how a support colleague assembles a kit. Do not observe private messages, sensitive screens or unrelated performance information merely because they are nearby.
Keep a simple evidence record:
| Stage | Observation to collect | Interpretation to investigate |
|---|---|---|
| Request | What information is entered and by whom? | Are requirements known at this point? |
| Preparation | How is a kit assembled and checked? | Is compatibility verified? |
| Collection | What can the employee try before leaving? | Does ticket closure reflect readiness? |
| First use | Which ordinary task is attempted? | Is another dependency missing? |
| Recovery | How is a problem reported and resolved? | Does the person know where to get help? |
The entries are research prompts for a fictional service. They do not assert that these problems occur in a real organization.
Build a map from evidence, not a preferred story
An experience map can organize research across stages, activities and difficulties. Distinguish observed events, participant accounts and team interpretations. If one employee waited a long time, record that occasion; do not immediately describe long waits as the normal experience of every employee.
Include people whose route differs. A remote employee, a night-shift worker and someone using assistive equipment may encounter dependencies that a standard daytime collection hides. The aim is useful coverage of situations, not a claim that a small interview group represents the entire workforce. Use the sampling guide when deciding which perspectives the inquiry needs.
Do not average conflicting experiences into a bland middle journey. Separate materially different routes. A smooth replacement for an existing employee may say little about a new colleague who lacks system access. Mark uncertain steps and return to them with the people who know the work.
Choose a change that addresses a mechanism
The fictional team might hypothesize that checking a complete kit with the employee before closing the collection ticket will expose incompatibilities earlier. That is a plausible proposal, not a proven remedy. It could also create a queue at the counter or delay colleagues who already have working equipment.
Use a service blueprint when the visible difficulty depends on preparation behind the counter. Identify who can check compatibility, where the required information comes from and who handles an exception. A friendlier message cannot compensate for an unavailable component.
Discuss the proposal with employees and service staff. Ask what would make it inconvenient, inaccessible or difficult to sustain. Give people a way to disagree without being labelled resistant. If a manager can overrule a concern, make the decision authority clear rather than implying that participation guarantees agreement.
Test the improvement during real work
Rehearse the proposed collection with a complete kit. Then plan a bounded, supported operational test if the rehearsal exposes no blocking issue. Keep an accessible recovery route and make clear that employees can obtain support regardless of their participation in the research.

Define what will count as useful evidence. Completion of the first intended task is more relevant than a cheerful collection interaction alone. Track practical side effects too: waiting, support effort and work transferred to managers. Measuring service benefits emphasizes relating measurement to intended benefits and an appropriate baseline.
Avoid a single success question such as “Was your experience good?” Ask what the person could do, what they could not do and what help they needed. A favourable rating may coexist with a failed task. If you use a survey, work through question design and the distinction between reliability and validity.
Do not promise attribution from a simple before-and-after comparison. Workload, staffing and the mix of equipment requests may change. A small test can show how the revised arrangement operates and what deserves a stronger evaluation; it cannot establish every claimed productivity benefit.
Make participation lead somewhere
Tell contributors what the team learned, what it changed and what remains unresolved. Remove identifying details when reporting incidents to a wider audience. An individual’s difficulty with a service should not silently become a performance judgement about that person.
Give the improved service an owner with the authority to coordinate its dependencies. Agree who maintains the kit specification, who handles an unavailable item and when the team reviews the arrangement. Otherwise a useful prototype can become another local workaround that disappears when a helpful employee changes roles.
Link the work to change readiness when implementation requires new capacity or authority. Use ADKAR for questions about an individual’s understanding and ability where appropriate, while continuing to investigate organizational obstacles. A missing adapter is not a lack of motivation.
Evidence limits and the next service review
Employee experience includes matters far beyond internal service design. This guide addresses a bounded operational problem; it does not provide a complete account of employment relationships, wellbeing or organizational culture. Institutional service-design guidance supports the inquiry, but it does not prove an effect for this fictional equipment service.
At the next review, bring one journey, the evidence behind its difficult step and a proposed change to the work behind it. Ask the employee and the service provider whether the new arrangement helps them complete the same task. Keep the unresolved dependency visible until someone with the necessary authority accepts it.
Sources:
- Government Digital Service (2026). Solve a whole problem for users.
- Government Digital Service (n.d.). Contextual research and observation.
- Government Digital Service (n.d.). Creating an experience map.
- Government Digital Service (n.d.). Measuring the benefits of your service.



