THE SHORT ANSWER

Trace where work waits, queues, repeats, returns or loses information. Compare demand with capacity, distinguish necessary judgment from routine handling, and measure the constraint before proposing automation.

Look for observable friction

Signal and question
SignalDiagnostic question
WaitingWhich queue or dependency blocks progress?
Repeated entryWhere is the same information typed twice?
ApprovalsWhich approval changes a decision?
HandoffsWhere is context lost between people or systems?
CommunicationWhich updates duplicate a reliable status?
Meetings/reportsWhich decision would fail without them?
Errors/reworkWhere does work return, and why?

Separate repetition from judgment

A repeated action may be automatable when inputs and rules are stable. A repeated decision may still require context, negotiation or accountability. Frequency alone is not enough.

  • How variable are the inputs?
  • Are rules explicit?
  • How costly is a wrong result?
  • Can exceptions be identified and routed?
  • Is reliable data available?

Improve the constraint first

Reducing ten seconds from an unconstrained step may not change cycle time. Prioritize the delay or quality failure that affects the outcome, then establish a baseline.

Continue to simplify before automating.

Apply the idea to one real workflow

Choose one current workflow. Record the present outcome, the proposed change, the accountable owner, the most important exception or failure, and one before-and-after measure. Test the smallest safe version before expanding it.

Sources & further reading

  1. Map and understand a user's whole problem

    GOV.UK Service Manual. Public-service guidance on mapping journeys, collaborating across boundaries and locating pain points. The module adapts the method to business workflows.

  2. Measuring the benefits of your service

    GOV.UK Service Manual. Guidance connecting time, cost and service measures to outcomes and benefits. It does not provide universal automation ROI assumptions.

Examples and exercises are illustrative unless attributed to a source. No independent expert review is claimed.

A correction, a counterexample or an experience worth sharing?

Join the conversation ↗