Stage gate decision on a major capital project business case

A public-sector infrastructure body's investment committee decides what to do with a capital business case at a stage gate: approve it, approve it with conditions, return it for rework, or reject it. ...

Published content

1. node: Context. The organisation is a public-sector infrastructure body delivering a five-year capital programme within a fixed envelope set by its sponsoring department. The proposal before the committee is the replacement of an operational maintenance depot serving three regional maintenance contracts, submitted at outline business case with a capital estimate of 47 million pounds. The submitted programme brings the replacement into operation in 24 months. Three other schemes in the programme are awaiting funding at the same gate, and the uncommitted headroom in the envelope is less than 47 million, so approving this proposal leaves too little for any of them.
2. node: At the outline business case gate, should the investment committee approve the depot replacement project as submitted, approve it subject to conditions, return it for rework, or reject it and release the funding to the portfolio?
3. node: Benefit. Approving as submitted preserves the programme and the spring construction start, and the strategic case for replacing the depot was accepted at the previous gate and has not been challenged.
4. node: Benefit. Conditions attach the committee's concerns to a verifiable event rather than to a hope: a project director appointed and design maturity at the standard, both testable before the next tranche.
5. node: Benefit. Returning the case is the response the gated model contemplates for a case whose five dimensions are unevenly developed: the deficiency is in the submission rather than in the proposal, and returning it addresses the deficiency directly.
6. node: Benefit. Rejection releases the envelope to schemes whose cases are complete, which serves the portfolio prioritisation the committee is responsible for.
7. node: Decision-maker. The investment committee, which holds delegated authority for capital above 10 million pounds. The sponsoring director presents the case and withdraws for the decision; the committee's decision is recorded with reasons and reported to the board.
8. node: Scope. The committee is deciding on this proposal at this gate. It is not deciding the merits of the depot strategy, which was settled at strategic outline case, nor the design, which is the sponsoring director's responsibility. Its options are those the gate provides: approve, approve with conditions, return, or reject.
9. node: Benefit. Returning the case keeps the proposal alive without reserving anything against the envelope, so the three competing schemes are not displaced by a proposal that has not yet demonstrated it can be delivered.
10. node: Consulted. The independent cost reviewer on design maturity; the programme office on the position of the three competing schemes; the property function on interim accommodation if the lease expires before a replacement is ready; and the three regional contract managers whose operations the depot serves.
11. node: Risk. Approving on the strength of the strategic case when the management case is weak is the characteristic failure of gated approval. The five cases are separate for a reason, and a strong case for doing something is not evidence of the ability to do it.
12. node: Risk. Conditional approval reserves the full 47 million against the envelope while releasing little cash, so it displaces three schemes for a project that may still not proceed. It is the option that most resembles having decided without having decided.
13. node: Risk. The strategic case for depot replacement was accepted at the previous gate and has not been withdrawn. Rejecting the submission rejects a proposal for its documentation rather than for its merits, and the need it addresses does not go away.
14. node: Action. Establish the cost of interim accommodation before the re-submission. It is absent from every option's figures and it is the consequence that distinguishes returning the case from approving it, so the committee is currently comparing options on an incomplete basis.
15. node: Constraint. The capital envelope for the five-year programme is fixed. Approving this proposal at the submitted value displaces other schemes; there is no option in which the money is found elsewhere.
16. node: Risk. Reserving the full 47 million against a fixed envelope displaces three schemes on an estimate the independent reviewer would not put a range around. The displacement is certain and the cost it is being incurred for is not.
17. node: Risk. Conditions are frequently reported as met on the papers rather than in substance, and the committee that set them is rarely the body that verifies them. Without a named verifier this option decays into unconditional approval with a delay.
18. node: Risk. The existing depot lease expires in 31 months. A rework that loses the construction season puts the replacement beyond that date and creates a cost for interim accommodation that has not been established and is in no option's figures.
19. node: Risk. The lease expiry becomes certain rather than at risk. Rejection makes interim accommodation a near certainty and its cost is still not established.
20. node: Action. Appoint a project director before re-submission, so that the re-submitted management case evidences delivery arrangements rather than describing them.
21. node: Constraint. The lease on the existing depot expires in 31 months. A replacement not operational by then requires interim accommodation, the cost of which is not in any option below because it has not been established.
22. node: Action. Re-appraise the two competing schemes appraised more than eighteen months ago against current construction cost inflation, so that the displacement comparison at the next gate rests on current figures rather than on stale ones.
23. node: Framework. The business case follows the five case model recommended by HM Treasury's Green Book: strategic, economic, commercial, financial and management cases. The model treats a business case as iterative across three stages corresponding to the approval stages, with detail added to each of the five dimensions at each stage. The committee's task at this gate is therefore whether the five cases are sufficiently developed for this stage, not whether the project is a good idea.
24. node: Framework. The Green Book requires costs to be valued at their opportunity cost: the value of the best alternative use of the resources. Its options appraisal compares ways of meeting a single objective, so comparing this proposal with competing schemes is the committee's portfolio task rather than something the business case performs. Within a fixed envelope, the opportunity cost of this proposal is the schemes it would displace, which is why displacement appears below as a criterion.
25. node: Review trigger. If the rework cannot be completed in time for the next gate, bring the matter back before the spring start is lost rather than after. The difference between a four-month delay and a fifteen-month one is a decision the committee should take knowingly.
26. node: Appraisal finding. The five cases in this submission are unevenly developed. The strategic case is strong and was accepted at the previous gate. The economic case is adequate. The management case is the weakest of the five: the delivery organisation is described as an intention rather than evidenced, and no project director has been appointed.
27. node: Review trigger. If interim accommodation in the lost-season case would cost materially more than the 2 million of additional spend that conditional approval commits, and the rework cannot be shown to hold the spring start, the balance changes and conditional approval becomes defensible. That figure is the one most capable of reversing this recommendation.
28. node: Assurance finding. The independent cost review records that 38 per cent of the design is fixed at this stage, against the 60 per cent the organisation's own assurance standard expects at outline business case. The estimate is described as indicative, and the reviewer declined to give a confidence range on it.
29. node: Review trigger. If the re-submitted case still does not reach the design maturity standard, the committee should reject rather than return a second time. A case returned twice is a case the sponsor cannot make.
30. node: Assumption. The three schemes displaced by this approval would each deliver returns comparable to those they were appraised at. Two were appraised more than eighteen months ago and neither has been re-tested against current construction cost inflation.
31. node: Assumption. A return for rework can be completed and re-submitted at the following gate without losing a construction season. The programme assumes a spring start; a re-submission that slips past that point costs a year rather than a quarter.
32. node: Approve as submitted
33. node: Approve subject to conditions, with the next tranche released only when the conditions are met
34. node: Return the case for rework and re-submission at the next gate
35. node: Reject and release the funding to the competing schemes
36. node: Recommendation. Return the case for rework and re-submission at the next gate, with the rework scoped to the management case and design maturity. It is the only option that neither commits the envelope against an estimate the independent reviewer would not put a range around, nor rejects a proposal whose strategic case the committee itself accepted. Conditional approval was considered and is not recommended: it reserves the full 47 million against the envelope while deciding nothing, which is the appearance of caution rather than caution.