Learning objective
Connect a measurable gap to tested causal mechanisms, countermeasures and follow-up without hiding uncertainty or side effects.
Why it matters
Concise reporting can create false coherence. A strong A3 remains short because the reasoning is disciplined, not because inconvenient evidence was removed.
Core concept
Every arrow should be challengeable. If a countermeasure cannot be traced to a cause, it is only an activity.
Scenario
Invoice corrections rise after a new intake route launches. A training campaign is proposed, but direct observation shows an ambiguous field mapping concentrated in that route.
Worked decision
Measure first-pass validity by channel, test the mapping correction, preserve a comparison path and monitor late payment and supplier burden. Training remains an option only where knowledge evidence supports it.
Decision rule
Prefer the smallest reversible countermeasure that tests the strongest causal mechanism while protecting affected users.
Ethical boundary
Do not turn employee telemetry into blame, hide workload shifts or expose supplier data. A3 ownership is accountability for learning, not personal fault.
AI lens
AI may organise authorised evidence and trace missing links. It cannot observe the work, establish causality or approve risk.
Quick check
After a change, error rate falls but supplier handling time doubles. What should the A3 show?
A. Only the improved error rate.
B. The balancing effect and a decision to adapt, stop or explicitly accept it.
C. A rewritten target.
D. No result until every metric improves.
Answer: B.
Next step
Complete the invoice-error case and test whether every countermeasure has a causal link and balancing measure.