Public agency fixed-price project portfolio
A public-sector digital services agency has 300 staff and 40 active projects. Average project duration is 18 months, and 70% of budget is consumed by change requests. Citizen satisfaction with online services is 58%, and call-center volume rose 20% last year. Procurement rules require fixed-price contracts and detailed specifications before work begins. Vendors deliver to spec, but users often reject the result. Security reviews and accessibility audits happen only at the end, causing late rework. No product owner role exists; project managers coordinate vendors. The new director wants agile at scale and a 12-month delivery target for all projects. The finance team measures success by budget spent, not outcomes. Two high-profile services are under ministerial scrutiny. The agency has a small in-house engineering group of 15 people, mostly maintainers. The director asks you to propose a realistic path.
1. Why is agile at scale unlikely to work in this agency as proposed? Identify the system conditions that would turn it into mini-waterfall.
Read the worked answer
Agile at scale without changing procurement, funding, and assurance will become mini-waterfall. Fixed-price contracts and detailed specifications lock scope before discovery. An 18-month average duration and 70% change-request spend show large batches and late learning. Security and accessibility audits at the end create rework instead of designing for compliance. Finance measures budget spent, not citizen outcomes. No product owner means vendors coordinate, not users. A 12-month target for all 40 projects is a top-down deadline that ignores discovery, risk, and the 15-person in-house engineering capacity. Agile ceremonies inside a fixed-scope contract cannot reduce uncertainty; they add meetings to a fixed plan. The system conditions are the contracting model, funding model, assurance timing, and capability mix. The director should change those conditions rather than mandate a method or a date.
2. Recommend a realistic path for the next 12 months. Include changes to procurement, assurance, and measurement, and say what you would pilot first.
Read the worked answer
Pilot one high-profile service with a cross-functional team, a product owner, and a discovery phase. Use modular contracting: a short discovery contract, then outcome-based delivery with exit points, not one fixed-price specification. Bring security and accessibility into design sprints and define acceptance criteria up front. Change finance to fund outcomes and capacity, with a portfolio that separates discovery, delivery, and run. Measure citizen satisfaction, completion rate, call deflection, lead time, and cost per transaction; stop using budget spent as the primary success metric. For the other projects, triage into sustain, reshape, or stop. Build in-house product and engineering capability by pairing staff with vendors. Do not promise 12 months for all 40 projects; commit to a pilot, a portfolio review, and transparent delivery confidence. Under ministerial scrutiny, publish user research and risk, not feature dates.