01 / Section
Start with the documented process
Write down the intended steps, handoffs, and outcome before reviewing captured work. Keep the description specific enough to compare: what should happen, in what order, and what marks a successful result?
Treat this document as a reference, not a verdict. It describes the process people are expected to follow; it does not establish what every attempt will look like.
02 / Section
Compare separate attempts, not one idealized story
Review evidence-linked attempts from captured batches one at a time. For each, note the recorded actions, visible handoffs, and connection to the intended business outcome. Keep the source evidence close to each observation so a reviewer can inspect where it came from.
Then compare distinct attempts. Look for steps that recur, steps that appear in a different order, repeated handoffs, or work that seems to take another route. A single attempt can show a variation; repeated patterns across attempts give a stronger basis for discussing how the workflow may operate.
03 / Section
Keep evidence, interpretation, and assumptions separate
Describe what the record shows before explaining what it might mean. For example: “The captured sequence includes a second review before submission” is an observation. “The process requires an extra approval” is an interpretation that may need confirmation.
A citation establishes provenance: it lets someone trace a statement to its source. It does not make the source complete or prove that an interpretation is certain. Captures can omit context, and a recorded sequence should not be treated as a complete account of all work.
04 / Section
Use app and site sequences as discovery signals
Keep app and site sequences distinct from evidence of a business outcome. They can suggest where to investigate—for example, a likely handoff or a possible return to an earlier step—but they do not, on their own, establish why someone switched tools or what work was completed.
Use those signals to form questions for process owners and other relevant sources. Confirm or revise the interpretation before changing the documented process.
05 / Section
Turn recurring differences into a review plan
Summarize the comparison as three separate things: what the procedure says, what the captured attempts show, and what remains uncertain. Focus discussion on workflow conditions and handoffs, not judgments about individual employees.
When a pattern appears meaningful, agree on what to verify next and who can confirm it. Update the procedure only when the evidence and operational context support the change; otherwise, record the open question and keep the original documentation intact.
