Strategic mandate and decision boundaries · Lesson 5

Challenge an investment request

6 min read

Lesson objectives

  • Challenge an investment case and calculate marginal cost while separating cash from released capacity.
Reading tools

When a polished investment request arrives, do not begin by asking whether its sponsor is trustworthy. Ask which decision is sought now and what must be known to make it. Funding a short discovery requires different evidence from committing to three years of delivery. This lesson turns a sponsor's presentation into a decision memorandum separating need from solution, assertion from evidence, and choice from implementation conditions. Do not penalise a team for disclosing uncertainty. Expose hidden assumptions and test their consequences instead. Governance should improve the choice, rather than produce a document that appears risk free or encourages the proposer to conceal inconvenient uncertainty.

The UK Green Book's Five Case Model helps examine a request from distinct perspectives: strategic fit, the value of alternatives, commercial arrangements, financial affordability and the ability to deliver and manage the work. It is a context specific public sector method whose questions can inform other settings, not a universal mandate. A project can serve strategy and provide value while being unaffordable this year. It can be affordable while lacking a capable supplier or an operational team willing and able to receive it. Do not fill five headings with the same promotional text. Each perspective asks a different question and requires information capable of changing the judgement. Where a material dimension is unresolved, limited discovery may be authorised instead of treating the full investment case as complete.

Examine the comparison baseline. Continuing without the project does not necessarily preserve current performance: a legacy system may deteriorate or become more expensive to operate. Equally, do not invent a disastrous baseline to flatter the preferred option. Develop a reasoned business as usual forecast, a feasible minimum option, and alternatives that solve the problem differently. The minimum identifies what is necessary to achieve the core purpose, while alternatives expose the value of additional scope. Two versions of the same product may overlook process redesign or buying an existing service. The aim is not to manufacture dozens of candidates. It is to ensure the range can test whether the sponsor's solution is genuinely the only reasonable route.

Separate cash savings from released time. Saving a thousand staff hours creates capacity for other work or better quality, but does not automatically reduce payroll. Avoid counting reduced outsourcing expenditure and the value of staff time doing the same work as separate benefits if they describe the same change. Put costs and benefits on the same time horizon, state whether prices are constant or changing, and identify exclusions. If present values are used, disclose the discount rate, rationale and timing of flows. A government rate should not silently become another organisation's rate. An initial undiscounted comparison is useful when its limits are explicit, but accurate addition does not by itself make it a complete investment appraisal.

After calculating, seek the strongest alternative explanation. A forecast saving may reflect a changing volume of requests. An early trial may overstate wider success because the most ready teams participated. Request evidence proportionate to the decision: a wider trial, cost challenge, supplier capability test or confirmation of operational commitment. Test the input most capable of changing the result. In the case above, if annual full option savings fall from 1.6 to 1.4 million, net cost becomes one million rather than 0.4 million. Do not merely label the case sensitive. Explain which threshold changes the choice and what better information would cost. Not every gap justifies a long delay. Compare the cost of waiting with the value of reducing uncertainty and the possibility of making a staged commitment.

End the memorandum with a specific recommendation, alternatives, reasons and conditions. State the authorised amount and period, who verifies conditions before expenditure, when the request returns, and what could stop it. Approval subject to completing all requirements is ineffective if no one knows the requirements or who certifies them. That wording transfers ambiguity into delivery. If the sponsor seeks a change to organisational objectives themselves, separate that choice from selecting a project to achieve existing objectives. Setting objectives is different from analysing routes to them. Preserve reasoned dissent in the record so later learning can examine what the board knew at the time, rather than reconstructing an apparently unanimous history after the result.

Lesson summary

An investment case distinguishes value, affordability and deliverability, ending with a choice and its conditions.