01 / Section
Start with what the record shows
When reviewing how business work happens, begin with the captured evidence: the records associated with an attempt, their order, and any citations that point back to their source. Shadow uses captured batches to reconstruct evidence-linked attempts at business outcomes.
A citation establishes provenance: it shows where a piece of information came from. It does not, by itself, establish that the information is complete, that the reconstruction is certain, or that the apparent sequence explains the outcome.
02 / Section
Treat interpretation as a separate step
A sequence of records may suggest that one action led to another, or that a handoff contributed to a result. Those are interpretations of the evidence, not the evidence itself. State them as such, and leave room for missing context or other explanations.
This distinction helps operations leaders discuss how a workflow may work without presenting a reconstruction as a definitive account of every decision or circumstance.
03 / Section
Compare attempts, not people
Comparing distinct attempts can reveal recurring workflow patterns: steps that appear in multiple attempts, common handoffs, or places where sequences differ. A repeated pattern is a useful signal for review, not proof that every case follows the same process or that the pattern caused a particular outcome.
Keep the unit of analysis on the work and its captured evidence. The purpose is to understand possible process patterns, not to assess individual employees or infer their productivity.
04 / Section
Keep discovery signals in their place
App and site sequences can offer clues about where a workflow may begin or which tools may be involved. Treat these as separate discovery signals, not as a complete account of a business attempt or proof of what someone intended.
When turning observations into a proposed process change, label planning assumptions clearly. That lets a team distinguish what the records show, what the team believes those records may mean, and what it proposes to try next.
