Observe the real workflow

Build an evidence-backed view of the work before estimating, automating, or accepting the inherited story.

Self-directed lesson

Field question
Who performs the work, what decision do they make, and what evidence changes that decision?
Working artifact
Field observation log
Effort
60–90 minutes

You will learn

  • Why sponsor description and operator practice are different evidence classes.
  • How a routine case and meaningful exception expose the real boundary.
  • How to preserve unknowns without turning assumptions into facts.

Field practice

  1. Find the process knower and observe one representative case.
  2. Record trigger, actors, systems, decision, exception, and exact source passage.
  3. Name the next evidence gap that could change the proposed boundary.

Before you begin

For live field evidence or a sanitized retrospective, bring an inherited brief, one workflow question, and permitted access to a person who performs or owns the work. For synthetic practice, use the supplied case pack and label the reconstruction as practice evidence.

Learning objectives

  • Separate observed behavior, authoritative policy, tested system behavior, and stakeholder assertion.
  • Map one representative case from trigger through decision, exception, and recorded outcome.
  • Name the smallest next field move that could materially change the proposed boundary.

Core lesson

What you need to know

01

Start with the decision, not the interface

A workflow is not a list of screens. It is a sequence of decisions made by people and systems under constraints. Begin by asking what arrives, who must decide, what evidence they inspect, what action follows, and where the result is recorded. This keeps the observation tied to work that can later be measured and designed.

Treat the sponsor, process knower, operating owner, and disposition authority as separate roles. One person may hold several roles, but do not infer one from another. A sponsor can describe intent; the person doing the work can reveal practice; policy can define what is allowed; only a scoped authority can change the engagement boundary.

Ask for one recent case, not a tour of the ideal process. Open with: ‘Could you reconstruct the last case that required judgment?’ Then ask what arrived, what made it difficult, which source changed the decision, what workaround was used, and what the next person received. Avoid questions that embed the sold solution or invite a yes-or-no description of policy.

02

Use claim-matched evidence

An observed representative case proves that something occurred in that case. It does not prove frequency or permission. A test against an exact system build proves technical behavior for that build. A policy is normative only within its stated scope and revision. A stakeholder assertion proves that the statement was made. Preserve those limits beside every consequential claim.

Observe one routine case and one meaningful exception when feasible. Record exact passages, source identity, date, scope, owner, and limitations. If access is denied or the case is not representative, mark the evidence gap explicitly and route a next action rather than filling it with an assumption.

03

Finish with a field move

A field observation is useful only when it changes what happens next. End with one of three movements: find or validate the person who knows or owns the work; observe a representative case; or resolve a cited conflict. Avoid converting every unknown into a backlog item. Pick the unresolved constraint most likely to change the outcome or boundary.

Before leaving, read back the path in neutral language, distinguish observation from interpretation, and ask the operator what you missed. Confirm what evidence may be retained, who may review it, and who owns the next move. The field-conversation kit provides a reusable outreach, observation, and readback script.

Worked field case

Claims-denial correction

A sold brief promises automatic corrected-claim resubmission. The FDE observes the billing specialist handling a denied inpatient claim.

Evidence available

  • Sponsor assertion: automation should remove manual submission work.
  • Observed case: the specialist checks a same-day encounter note and restores modifier 25 before submission.
  • Policy passage: only an authenticated specialist may approve and submit a corrected claim.

Reasoning path

  1. Map denial receipt, evidence gathering, correction recommendation, specialist approval, submission, and claims-system readback.
  2. Mark the automatic-submission claim contested because observed practice and policy narrow the allowed action.
  3. Choose the next move: observe a second representative exception and confirm the policy owner before reframing scope.
Result

The working proposal becomes recommendation-only pending a scoped human disposition; observation alone does not change the authorized boundary, and the original promise remains preserved.

Practice exercise

Observe or reconstruct one case

For live field evidence, observe a recent consequential case with the operator. For a sanitized retrospective, reconstruct the case from preserved sources. For synthetic practice, use S3 from the case pack and state that it demonstrates reasoning—not field observation or customer authority.

  1. Name the evidence mode. Agree the observation and evidence-handling boundaries, then observe or reconstruct the trigger, actor, decision, input evidence, action, exception, and system of record.
  2. Classify each consequential claim by evidence basis and state its limitation.
  3. Record one routine path and one exception or explicitly state why the exception could not be observed.
  4. Name the unresolved constraint and one next field move.
Keep

A source-linked field observation log another reviewer can distinguish from sponsor narrative or design intent.

Review your work

Field rubric

  • The case is bounded, dated, source-linked, and labeled as live field evidence, sanitized retrospective, or synthetic practice.
  • Observed behavior, assertion, policy, and tested behavior are not conflated.
  • Unknowns and limitations remain explicit.
  • The next field move is singular, owned when known, and consequential.

Complete when

  • One case is mapped end to end; any representativeness claim is evidenced, and synthetic work is explicitly labeled as practice.
  • Every consequential claim is source-linked or explicit unknown.
  • The next field move can be acted on without interpreting the log.

Check your understanding

Could another reviewer distinguish what you observed, what someone asserted, and what remains unknown?

Working formats

Use the editable Word workbook for field-facing work. The source Markdown remains useful for technical records and versionable reference.

Guided study and progress

Open this mission in the guided school to confirm the exercise and rubric, save completion in this browser, and resume the ten-mission path.

Open guided mission

Compare and go deeper

This lesson is self-contained. Compare your work with the calibrated answer, then use the attributed references when you need governed detail or additional implementation practice.

Annotated answerFDE Guide reference ↗Attributed technical depth ↗