A Center of Excellence Under Review
The Department of Transport Services has run a BPM program for three years. A central BPM Center of Excellence of nine analysts supports 12 divisions. Only three divisions have active process-improvement projects; the others say the CoE is slow and does not understand their work. The program dashboard reports 214 workshops delivered, 46 process models published, and 98% training attendance. The minister's office wants evidence of service improvement: licence renewal time has fallen from 21 to 19 days against a target of 10 days. Two divisions have hired their own Lean consultants, duplicating effort and creating incompatible methods. The CoE budget is €1.5 million and under review. The CoE has no service catalogue or chargeback model. The CIO proposes decentralizing all nine analysts into divisions. The program manager must present options to the executive board.
1. Diagnose the program: distinguish symptoms from root causes and assess its maturity.
Read the worked answer
Symptoms: low demand (3 of 12 divisions), slow service perception, activity-based reporting, duplicated consultants, and a 19-day cycle time against a 10-day target. These are visible but not causes.
Root causes, in order of leverage:
- Measurement is supply-side. Workshops, models, and training attendance measure CoE activity, not process outcomes. The program cannot prove value, so budget review becomes a popularity contest.
- No demand management or service catalogue. Divisions do not know what to ask for, how long it takes, or what it costs. The CoE appears slow because expectations are undefined.
- Weak process ownership. If divisional executives do not own process KPIs, improvement is a CoE project rather than a business change. Only three divisions engaging is a symptom of that.
- Mandate and operating model are unclear. A central team of nine cannot serve 12 divisions at high intensity; it is spread thin and cannot specialize.
- Method fragmentation. Divisions hiring Lean consultants shows unmet demand and absence of an enterprise method.
Maturity: likely level 2, repeatable but not managed. Models exist and projects happen, but there is no portfolio, no standard method, and no outcome measurement. The board should not read this as failure of BPM; it is failure of program design.
2. Evaluate the CIO's decentralization proposal and recommend an operating model.
Read the worked answer
Decentralization solves responsiveness but creates three risks: loss of scarce method skills, inconsistent process architecture, and duplicated tools and consultants. Full centralization caused the current bottleneck. The better fit is a hybrid.
Recommended model:
- A small central core of three to four people owns method, process architecture, tools, standards, training, and the portfolio. It runs a service catalogue with service levels.
- Embedded BPM analysts, the other five to six people, sit in clusters of divisions, report professionally to the central core and operationally to divisional leadership. This gives proximity without fragmentation.
- Divisional process owners are accountable for outcomes; the CoE does not own divisional processes.
- A demand-intake and prioritization process with published criteria, so the CoE is not a first-come queue.
- A community of practice and an approved supplier panel for Lean and other methods, so divisions can buy capacity without creating incompatible approaches.
Decision criteria to present: speed of response, method consistency, cost, capability retention, and scalability. The hybrid scores better than either pure option on all five, provided the service catalogue and SLAs are real. If the board insists on decentralization, keep the architecture and method functions central, or the duplication will worsen.
3. Redesign measurement and benefits realization for the program.
Read the worked answer
Move from activity to outcome with a causal chain. Start with the ministerial target: licence renewal from 21 to 10 days. Then define a benefits dependency network: program deliverables to process changes to operational KPIs to citizen outcome to ministerial goal.
Operational KPIs: end-to-end cycle time (median and 90th percentile), touch time, handoffs, rework rate, first-time-right rate, backlog age, and digital completion rate. Add adoption and capability indicators as leading measures: number of divisions with an active owner, percentage of process changes with a measured baseline, and time from demand intake to first delivered change.
Governance: each benefit has an owner, a baseline, a target, a measurement method, and a review date. Report a balanced scorecard quarterly covering outcomes, portfolio health, capability, and customer satisfaction. Stop reporting workshop counts as a headline metric; use them only as a diagnostic.
Benefits realization: stage-gate funding. Release tranches only when the previous tranche has produced measured change. Conduct post-implementation reviews at 3 and 12 months. If a projected benefit does not materialize, document the learning and reallocate. The board's test should be whether €1.5 million can be traced to a measured improvement in a service a citizen experiences. Until that sentence is true, the program is not managed.