LEARN. PRACTICE. GROW.1. Connect improvement to outcome
A fictional APS team installed batch automation. Installation is a deliverable; the objective was fewer error-prone manual interventions. Define which tasks are measured, how an error is recognized, the baseline and the observation period. Add a quality boundary so reduced effort does not increase incorrect files. An objective such as “greater efficiency” does not clarify what to decide if execution minutes fall while returns rise. Direction should explain the intended outcome and acceptable boundaries; planning translates that into activities, capacity and evidence.
2. Admissibility before scoring
A score helps compare proposals within agreed conditions. It does not authorize bypassing a mandatory control. In this workshop’s model, A scores 12 and uses three days but fails the access control. B scores nine and uses three; C scores five and uses two. With five available days and independent B/C, B+C totals 14 points and is the best admissible portfolio. A+C would total 17 but remains excluded while the control is missing. Points are a local learning rule, not financial value or a universal prioritization recommendation. Record conditions and the decision for later review.
3. Capacity shared with RUN
Two improvements may fit individually without fitting together. With 16 person-days and 12 reserved for RUN, four remain. Two three-day proposals need six and create a two-day shortfall. Do not count the same person twice or turn total availability into capacity for any skill. Discuss sequence, scope or suitable capacity with authorized decision-makers. In the portfolio case, five days are already net of reserves; subtracting RUN again would distort the model. In real operations, a reservation is an assumption to review against incidents and absences, not a guarantee of future availability.
4. Authority has scope and validity
A 10,000-euro delegation for laboratory work only does not authorize a 9,000-euro production intervention. Amount is one condition and scope is another. Likewise, pilot authorization for one application does not automatically cover five applications with different dependencies. Confirm who decides expansion, what risk changed and which recovery capability needs demonstration. An exception ending Monday at 18:00 does not remain valid Tuesday merely because work is ongoing. Follow local escalation and operating conditions. Do not retrospectively rewrite dates or assume renewal from participants’ silence.
5. Compare effort over time
A 120-hour automation saves six net hours monthly for two months and 12 thereafter. The first two recover 12 hours; 108 remain, requiring nine months. Effort break-even occurs in month 11 under fixed rates. Using 12 from the start would yield ten months and ignore adaptation. Define whether savings already deduct maintenance to avoid subtracting it twice. An hours-based calculation does not demonstrate profit, financial return or lower spending. Use it to compare effort and released capacity while keeping other relevant dimensions and assumption uncertainty separate.
6. Workshop and summary
Prepare a portfolio decision using A, B and C and explain why a higher-scoring combination may be inadmissible. Then use a 96-hour automation: months 1–2 save four net hours, months 3–8 eight and subsequent months five. By month 12, 76 accumulate; by month 16, 96. Show the forecast and conditions that may change it. Summary: connect work to outcomes, apply mandatory criteria, respect capacity and authority, and compare effort using time-consistent rates. A decision is incomplete without ownership, next steps and expected evidence.
Initial effort: 96 h
Months 1-2: 2 x 4 = 8 h
Months 3-8: 6 x 8 = 48 h
Months 9-12: 4 x 5 = 20 h
Cumulative at month 12: 76 h
Remaining: 20 h / 5 h per month = 4 months
Effort break-even: month 16Net capacity: 5 days. A: 3 days/12 points, inadmissible. B: 3/9. C: 2/5. B+C uses 5 days and totals 14 admissible points.
Common pitfalls
Score as control waiver; double-counted capacity; budget as the only authority condition; constant savings assumed without evidence.
Related topics: Governance and authority · Continual improvement
A sound proposal needs a useful outcome, admissible conditions and actual means to execute and verify.
Reference: ITIL 4 Strategist: Direct, Plan and Improve · ITIL 4 DPI; observed JA v1.3.1, current detailed revision comparison pending