The Discount That Arrives Before the Design Gate
Riverbend School District has 14 months and a fixed £2.4m budget to replace its student records system. The vendor offers a 15% licence discount if the contract is signed within six weeks, which is before the requirements phase ends. The sponsor wants development to start immediately to 'save eight weeks'. The steering committee meets monthly. Two deliverables are non-negotiable: statutory safeguarding reporting must be in place by the start of the academic year, and historical records must be migrated with a full audit trail. The project manager has a team of nine, none of whom have done a data migration at this scale, and the previous two IT projects at the district overran because users were not involved until testing.
1. Should the project manager recommend starting development before requirements are signed off? Justify your position with reference to the phase gate and the discount deadline.
Read the worked answer
No, though the deadline should still be pursued. The 15% licence discount is a procurement issue, not a build issue: it can be secured by signing a licence agreement whose obligations are conditional on requirements sign-off, or by paying a short option fee, without authorising development. The phase gate should not be passed because its exit criteria are unmet: there is no signed requirements baseline, no agreed acceptance criteria for the safeguarding reporting, and no data migration approach despite the team's inexperience at this scale. The cost of change rises steeply once design is frozen, and safeguarding reporting is statutory, so rework there carries regulatory risk, not just cost. Starting build early would also let the eight-week saving be consumed invisibly by rework.
The useful compromise is fast-tracking only low-irreversibility work in parallel: environment provisioning, data profiling on a sample extract, and procurement. These generate knowledge without freezing design decisions.
2. The sponsor asks whether the project should be run as a predictive (waterfall) or an adaptive (agile) life cycle. Recommend an approach and explain what determines the choice.
Read the worked answer
Recommend a hybrid life cycle, and justify it by the type of uncertainty in each part of the work rather than by preference.
- Predictive elements: the statutory safeguarding reporting and the historical data migration have fixed, externally audited acceptance criteria and a hard regulatory date. These suit a defined, documented, gated approach with formal sign-off.
- Adaptive elements: the user-facing workflows and the school-office screens are where the district's previous failures occurred, because users were only involved at testing. Short iterations with real records-office staff would surface misunderstanding early, when it is cheap.
- Deciding factors to state explicitly: stability of requirements, regulatory and contractual constraints, frequency of stakeholder availability, team experience (nine people with no migration track record argues for more, not less, structure), and how the contract prices change.
The life cycle choice must also be reflected in governance: a single requirements gate can still authorise iterative delivery inside a release, provided the statutory deliverables have their own fixed baselines.
3. Define the exit criteria the project manager should propose for the requirements phase.
Read the worked answer
Exit criteria are the evidence that lets a gate be a real decision rather than a formality. A defensible set would be:
- A requirements baseline signed by the records office, the safeguarding lead and the finance owner, with each requirement traceable to its source.
- Documented acceptance criteria for the safeguarding reporting that cite the applicable statutory specification and the evidence an auditor would expect.
- An approved data migration strategy, validated by a trial migration of a representative sample extract with the audit trail demonstrated end to end.
- A costed, scheduled plan for the next phase consistent with the £2.4m envelope and the 14-month end date, showing the float available.
- Stakeholder engagement commitments named and dated, addressing the previous projects' failure to involve users before testing.
- A risk register reviewed at the gate, with the top three risks owned and responded to.
Each criterion should be objectively assessable, so the committee can genuinely decide to proceed, proceed with conditions, or hold.