Cover photo: Cht Gsml / Unsplash. Illustrative photograph, not a client project.
Start with the decision being requested
A programme business case should explain why a change deserves investment, which option is preferable and what would make that judgement wrong. If the only credible outcome of reading it is approval, the options probably need more work.
Write the decision on the first page: the next commitment, the amount, its limits and the person authorised to make it. Approval to investigate an option is different from approval to procure it or roll it out across every location.
A useful business case can support a decision to stop. It should also show what happens if the organisation does nothing beyond maintaining its existing service.
Use the current guidance without importing its jurisdiction
HM Treasury updated its business-case guidance on 30 June 2026, combining the previous project and programme guides. Use that edition rather than an archived 2018 programme guide.
The 2026 guide uses five connected perspectives: strategic need, economic value, commercial arrangements, financial affordability and management/delivery. Chapter 5 addresses programme business cases. These UK public-sector requirements do not become UAE legal requirements because a team adopts the structure.
For the practical worksheet here, ask five questions: what needs to improve; which options could achieve it; how would each be sourced; can the organisation fund it; and can it deliver and operate it? This is a starting aid, not a compliant Treasury submission or a substitute for the appraisal your investment authority requires.
Compare three plausible options
Consider a fictional service operation choosing how to reduce manual case handling. All figures below are assumptions in GBP thousands, at constant prices, over three operating years. Setup is counted once; annual operating cost is counted three times. There is no discounting, tax, inflation, residual value or monetised benefit in this simple cost screen.
| Option | Setup | Annual operation | Three-year cost | Delivery assumption |
|---|---|---|---|---|
| Maintain current service | 0 | 180 | 540 | Existing staff and process |
| Redesign process and configure current tools | 60 | 135 | 465 | Process owner and four-month delivery window |
| Replace platform and redesign process | 150 | 105 | 465 | Procurement, migration and nine-month delivery window |
The second option costs 60 + 3 × 135 = 465. The third costs 150 + 3 × 105 = 465. Both are 75 below the baseline across this horizon, on these assumptions. That does not make them equally attractive, and it is not a return-on-investment calculation.
The replacement may offer capability the current tools cannot provide. It also requires more upfront cash and a longer transition. Test whether the desired service outcome genuinely needs replacement before treating the larger intervention as inevitable. Keep accessibility, service quality and other non-monetised consequences visible rather than forcing invented prices onto them.
Test the assumption that chooses the winner
If replacement running costs are 20% higher than assumed, annual cost becomes 126 and its three-year total becomes 528. The difference from the baseline narrows to 12. If the configuration option’s setup cost rises from 60 to 90, its total becomes 495. These are separate sensitivity tests, not forecasts or probability estimates.
Next test timing. A delayed launch may require the old and new service to operate together. Price that overlap explicitly, including staff cover, migration, training and supplier exit. The simplified table excludes those costs, so it is not sufficient for an investment decision without adding them.
Use ranges backed by evidence. A supplier quote, an internal capacity estimate and a hoped-for saving have different levels of confidence. Record the source, date and accountable reviewer for each material assumption. For a UK/UAE programme, state the currency and any exchange-rate assumption; do not add pounds and dirhams in the same total.
Make a bounded recommendation
A defensible next step in this fictional example could be a short feasibility assessment of the configuration option, with a spending cap and an explicit test of the missing capabilities. It would not yet be authority for a full rollout.
Record the preferred option, why others were rejected, funding source, commercial route, delivery capacity and conditions before the next commitment. Identify dependencies on operational managers who must release staff. An unfilled project plan does not become affordable because the supplier price is low.
Give the case a review trigger: a failed feasibility test, a cost threshold, a material schedule change or evidence that the original need has changed. Once investment is justified, use a benefits realisation plan to assign ownership and test whether the expected outcomes arrive. Do not quietly turn forecast value into reported savings.
Practical questions
Is the cheapest option always the preferred option?
No. An option must meet the required outcome and constraints. Compare cost alongside service quality, risk, delivery feasibility and consequences that cannot sensibly be monetised. A low-cost option that cannot meet the need should not win by arithmetic.
Can a programme use one case for several projects?
The programme case can explain the combined investment rationale and dependencies. It does not remove any project-level approvals your organisation requires. Identify what each approval authorises and avoid counting the same benefit once per contributing project.
Should sunk costs justify continuing?
Money already spent is not recovered by spending more. Show past expenditure transparently, then compare the remaining costs, benefits and consequences of continuing, changing course or stopping. Include unavoidable exit commitments in the forward-looking comparison.
Use the worksheet
Download the editable worksheet (CSV) ↓
Open the CSV in your spreadsheet application, replace the illustrative entries and add rows for your programme. It contains plain values, no formulas or macros. Keep evidence references in the sheet and sensitive supporting material in your own controlled systems. Nothing is submitted through this download.
Reviewed 8 November 2026. Examples, figures and working methods are illustrative editorial material, not measured client results, professional certification or a substitute for your organisation’s required approvals.

