Skip to content

Analysis Protocol

The core method. Run it on one work unit: a role, process, task or system. Output goes into a case file (template).

Before starting

  • Name the work unit precisely.
  • Name the reference system (team, organization, customer base…). Waste and value depend on it.
  • State what you already believe about it (the default view). Usefulness is measured against this.

Steps

1. FLOW        Which flow does it touch?
               Primary function: access / convert / maintain / steer / allocate / explore
               Secondary functions:

2. CONSTRAINT  What constraint does it build, maintain or use?
               What would the flow do without it?

3. LEVERAGE    What information does it hold or apply?
               Small effort → how large a flow? (rough ratio or qualitative: low/med/high)

4. SELECTOR    Who decides it continues? What do they actually measure?
               Gap between metric and real flow?
               Speed and accuracy of that feedback?

5. VARIATION   How do new ways of doing it appear?
               What does trying a variant cost? Who is allowed to try?

6. WASTE       Deletion test: what decays if it stops, how fast, who notices first?
               Verdict: needed / buffer / vestigial / self-serving / search cost

7. EXPOSURE    Which of its functions is getting cheaper (AI, automation, tooling)?
               Where does scarcity move?

8. LEVERS      What can change? (see Change Levers)
               flow · constraint · information · selector · variation

9. IMPACT      Per level, for the chosen lever:
               result / team / org / customers / economy
               effect on flow · sign (+/–) · delay · who notices

Quick version (5 minutes)

  1. Primary function?
  2. Who selects it, on what metric?
  3. What decays if it stops, and who notices first?
  4. Which part is getting cheap?
  5. One lever and its effect on the next level up.

Output check

  • Did any step produce something you didn't already believe? If not, note it: either the default view is right or the step is weak.
  • Which steps were guesses? Mark them.

Known weak steps

Steps 3 (leverage) and 9 (impact) are the least operational. Expect them to change after the first cases.