A business case is the document that argues for a project and answers why this project, why now and why this option among the others. It is an argument about the use of money that could have gone somewhere else, which is why it survives long after the approval meeting it was written for. PMI treats the business case as the sponsor's document and as the input the charter draws on, and the examination tests it through the financial measures and through scenarios in which the case has stopped holding.
The sections below cover the options a case compares, the three financial measures and what each one hides, a worked comparison of two options, when the case is revisited and why a project running well can still be cancelled.
The options, including doing nothing
A case with one option in it is a request. The comparison is what makes it checkable, so a serious case sets the proposal against at least a few alternatives. Buying where building was assumed, doing a smaller version, deferring for a year and solving the problem with a process change are all real options that get skipped because somebody had already decided.
Doing nothing belongs in every case, and it is the option most often left out. It is rarely free. The costs of carrying on as before include the growing manual workload, the risk that eventually lands and the revenue that goes elsewhere. Writing those down does two useful things. It sets the baseline that every other option is measured against, and it occasionally reveals that the honest answer is to leave the situation alone.
Payback, net present value and internal rate of return
Each option is then measured, and the examination gives the figures. What it tests is what each measure rewards and what it hides, because the three routinely rank the same two options in different orders.
| Measure | What it answers | How to read the result |
|---|---|---|
| Payback period | How long until the cumulative benefits cover the cost | Shorter is better. It is a reading of exposure, with no view of value, it counts nothing that happens after the payback point, and it treats money arriving in year five as worth the same as money arriving now |
| Net present value | What the whole stream of money in and out is worth in today's terms at the chosen discount rate | Positive means the investment beats the discount rate and is worth making on its own terms. The answer is an amount, so a large project can win on it while being much worse per pound invested |
| Internal rate of return | Which discount rate would bring net present value to zero | Higher is better, and because it is a percentage it compares investments of different sizes. It is blind to scale, so a small project returning forty per cent can create less value than a large one returning fifteen |
Payback period is how long until the benefits cover the cost. It is easy to explain and everyone understands it, which is why it survives. It is read as a measure of exposure, because a short payback means less time for the world to change underneath the assumptions.
Net present value adds up the money in and out across the life of the investment, discounting later amounts because a pound in five years is worth less than a pound now. The discount rate carries the organisation's cost of capital and its appetite for waiting, so raising it punishes benefits that arrive late and quietly favours short projects.
Internal rate of return is the discount rate at which net present value would come out at zero. Because it is a percentage it compares projects of different sizes, which is why portfolio committees like it, and that same property is what makes it silent about how much money is actually at stake.
Two options worked through
Higher net present value, higher internal rate of return and shorter payback are all preferable, so the three measures ought to point the same way. They disagree more often than people expect, and the disagreement is the part worth understanding.
Take two proposals that each cost 400,000 pounds, spent at the start, appraised over five years at a discount rate of ten per cent. Option A replaces an ageing settlement engine, where the saving arrives immediately and then tails off as the old volumes fall away. Option B rebuilds the customer portal, where adoption ramps slowly and then holds.
| Year | Discount factor at ten per cent | Option A benefit | Option A discounted | Option B benefit | Option B discounted |
|---|---|---|---|---|---|
| 1 | 0.909 | 200,000 | 181,800 | 60,000 | 54,540 |
| 2 | 0.826 | 200,000 | 165,200 | 120,000 | 99,120 |
| 3 | 0.751 | 100,000 | 75,100 | 220,000 | 165,220 |
| 4 | 0.683 | 50,000 | 34,150 | 260,000 | 177,580 |
| 5 | 0.621 | 50,000 | 31,050 | 260,000 | 161,460 |
| Total | 600,000 | 487,300 | 920,000 | 657,920 |
Option A recovers its 400,000 by the end of year two. Option B takes until the end of year three, because its first two years bring in only 180,000 between them.
| Measure | Option A | Option B |
|---|---|---|
| Payback period | 2.0 years | 3.0 years |
| Net present value at ten per cent | 87,300 | 257,920 |
| Internal rate of return | About 21 per cent | About 28 per cent |
Payback picks Option A. Net present value and internal rate of return both pick Option B by a wide margin, and all three readings are arithmetically correct. Payback is telling the truth about the thing it measures, which is how long the money is exposed, and it stops counting at the moment the cost is covered. Everything Option B earns in years four and five, which is more than half of what it earns in total, sits outside its field of view.
The reading to take into a scenario question is that payback ranks options by risk and net present value ranks them by value. Where the two disagree, the tie is broken by how much confidence anybody has in the later years. Where the market in year four is genuinely unknowable, Option A's short exposure is worth paying for. Where the later figures are solid, choosing A over B gives away 170,620 pounds of value to buy a year of certainty, and that trade is a decision for the sponsor and never a property of the spreadsheet.
The interesting work is never the arithmetic. It is whether the benefit figures were agreed with the people who will have to deliver them, and how the costs were estimated.
When the case is revisited
Those benefit figures rest on assumptions about the world, and assumptions go stale. The case is therefore reopened at each stage gate, whenever a significant assumption is broken and whenever the cost forecast moves outside tolerance. What gets rechecked is narrow and specific, meaning whether the benefits are still wanted, whether the cost to complete has moved and whether the option chosen is still the best one now that some uncertainty has resolved.
Sponsors own this. A case rechecked by the project manager alone is a project judging itself, and the answer is predictable.
A project can be cancelled while executing perfectly
Sometimes the recheck fails, and the hardest thing a governance body does is stop a project that is running well. It happens when a competitor ships first, when a regulation removes the market, when a reorganisation delivers the benefit another way or when the cost to complete rises past what the remaining benefits justify.
All of that is a judgement about the world and says nothing about the team. The test at any point is the same one the original case asked, which is whether the money still to be spent buys more here than anywhere else. Money already spent belongs to the past and stays out of that calculation, however uncomfortable it is to write off. Organisations that cannot stop projects end up unable to start them, because their capacity is permanently committed to decisions taken years ago.