Scaling a Successful Pilot Without Losing It
After an 18-month pilot, the consumer division of a 5,000-person retailer has four agile teams. Lead time fell from 45 to 12 days, employee engagement rose 14 points, and customer satisfaction for the pilot journey rose 9 points. The board now wants the same model across three divisions and five countries within a year. There are 12 new teams, but only three experienced coaches. The pilot's success depended on a co-located product owner, a supportive VP who protected the teams from ad hoc requests, and a funding model that let the division reallocate €1.5 million without central approval. In the new divisions, middle managers have been told their roles will change but no details exist. A central PMO has issued a standard agile maturity dashboard with 24 metrics, including velocity and utilization. Two pilot team members have been asked to return to their old roles because their managers need headcount. The transformation lead has a budget for six months and is under pressure to show quick wins.
1. Which conditions made the pilot work, and which should be treated as non-negotiable when expanding? Distinguish principles from practices.
Read the worked answer
Pilot conditions: four teams, a co-located product owner, a protective VP, and €1.5 million of delegated reallocation. The outcomes-lead time 45 to 12 days, engagement plus 14, customer satisfaction plus 9-are the evidence. Non-negotiable principles: clear product ownership with real priority authority; funding decision rights at the value-stream level; protected capacity from ad hoc requests; leadership sponsorship that removes impediments; and teams empowered to improve their own system. These are causal conditions, not ceremony choices. Adaptable practices: co-location, which can be replaced by deliberate communication; specific ceremonies; tooling; and team size. The expansion plan should not copy the pilot calendar; it should replicate the decision rights and sponsorship. Risks: three coaches for 12 teams is thin; middle managers have no defined role; the 24-metric dashboard will drive compliance; pulling pilot members back destroys the social network that made the pilot work. The board's one-year, five-country target is a portfolio bet, not a rollout schedule. I would test readiness by value stream, not by country.
2. Design a 12-month expansion with sequencing, guardrails, and metrics. Address the six-month budget pressure.
Read the worked answer
Wave 1, months 0 to 3: one division, one value stream, two teams, plus a seed pair from the pilot. Do not start 12 teams. Appoint a product owner with real authority and a business sponsor; define a delegated funding envelope; agree guardrails on risk, compliance, and architecture. Wave 2, months 3 to 6: add two value streams where sponsors are ready; create internal coaches by pairing the three experienced coaches with six internal candidates. Wave 3, months 6 to 12: expand only if flow and engagement metrics hold.
Middle managers: define their new accountability-capability building, removing impediments, connecting strategy, and managing external dependencies-not status reporting. Metrics: lead time, decision latency, change-failure rate, customer outcome, engagement, and flow efficiency. Avoid comparative velocity and utilization. Six-month budget: spend on coaching capacity and manager development, not tooling; use the pilot outcomes to secure the next tranche. Quick wins: publish decision-latency reduction and one customer outcome, not ceremony counts.
3. The PMO issues a 24-metric maturity dashboard and two pilot staff are being pulled back. What should the transformation lead do?
Read the worked answer
On the dashboard: do not refuse measurement; negotiate the smallest set that answers whether the system is improving. Keep five to seven metrics: lead time, decision latency, change-failure rate, customer outcome, engagement, and flow efficiency. Treat maturity levels as diagnostic, not targets. Ask the PMO to separate mandatory financial and regulatory metrics from transformation metrics and to stop comparing velocity across teams. Offer a 90-day trial and review.
On staff: losing pilot members is a real risk to the social network. Negotiate a time-boxed mentoring role-for example 20% for six months-with explicit knowledge transfer: pairing, decision-rights coaching, and documentation of the value-stream setup. If managers insist on headcount return, ask them to fund a replacement coach or accept a slower expansion in their area. Escalate with the board sponsor if needed. The principle: protect the conditions that produced the results, not the individuals forever. Build a coaching bench so the expansion does not depend on two heroes.