A decision-maker has little time and a fair suspicion of marketing reports. A performance review for a business case is a short, honest account in four parts: what we did, what happened, what we learned, and what it means for the next decision. Read the worked example later in this topic first.

Start from your own numbers. You built the pieces in Module 6 and Module 14. Pull them together: the objective and the baseline you were measured against, the result, the cost, the ROI, and the assumption it rests on. If part of the result would have happened anyway, say how you estimated that and that it is an estimate.

Be honest about misses. A missed target does not sink a case. A hidden one does, because the first awkward question will find it. Say what was missed, by how much, and what you think caused it. Then say whether the investment still earned its money. Both can be true.

End on meaning. Finish with one sentence beginning "So what this tells us is". That is the bridge to the options.

Keep it short. If the review runs past one page, the reader is being asked to do your analysis for you.

A business case sets out why an investment is needed, the options, the costs, the benefits and risks, and a recommendation. It is an argument for a decision: what we should do, why, what it costs, what could go wrong and who will do it.

The Five Case Model. HM Treasury's framework splits the argument into five parts:

UK example: the Treasury's guidance. HM Treasury publishes Guidance on Developing Business Cases for projects and programmes, based on the Five Case Model (first published 16 May 2024). It sits beside the Green Book, the Treasury's appraisal guidance, which has a 2026 edition after a 2025 review. Both are written for public spending, so borrow the structure, not the rules. Check gov.uk for the current versions and for how the guidance handles options.

Options appraisal. Compare realistic options, and always include doing the minimum. It is the baseline: every other option has to beat it. An option with no real alternative beside it is a decision already made.

Use your milestones. Your options should come from work you have already done. If you planned a second market in Module 13, that plan is a ready-made option, with its contribution per customer. The lifetime value from Module 12 turns customers into money, and the campaign you planned in Module 7 is what your review judges.

Judging the money. ROI is (benefit − cost) ÷ cost. Payback is how long until the money back covers the cost. Both need a stated benefit, and the benefit needs a stated assumption.

Sensitivity. Change the assumption you trust least and see what happens. Find the break-even point, the result where benefit equals cost, and ask how far the plan is from it.

A risk register is a list of what could go wrong, each with its likelihood, impact, owner and the action to reduce it. It is the part of a case that shows you have thought past the good news.

Look in five places. Commercial risks affect sales and money, such as slow uptake or a currency move. Operational risks affect delivery, such as capacity or delays. Legal risks come from rules, such as consumer and data law. Ethical risks are about doing the right thing, such as claims you can't support. Reputational risks are what the others cost you in trust. You met the rules side in Module 5, and an unsupported claim is both an ethical and a reputational risk.

Score and rank. Give each risk a likelihood and an impact from 1 to 5, and multiply. A risk scoring 4 for likelihood and 3 for impact scores 12. The scale is a convention, not a law of nature, and the scores are your judgement. Show your reasoning so someone can disagree with it.

Every risk needs an owner and an action. An owner is one named person. An action says what will be done and by when, or what would trigger it. "Monitor" is not an action.

Plan the trigger. For the biggest risks, decide in advance what you'll see that tells you to act or to stop.

Risks you choose to accept. Some risks aren't worth reducing. Write down which, and why.

The best case still fails if the people who must act on it aren't with you. Leading marketing means getting a decision made and then getting the work done, often by people you don't manage.

Know what each person needs. The finance adviser wants to see the sums and the downside. The managing director wants a clear choice and a recommendation. The delivery team wants to know what changes for them. An agency wants a clear brief, as you learned in Module 10. Write down, for each, what they need from you and what they might worry about.

Lead people you don't manage. You have no authority over an agency or another department. You have clarity, fairness and your word. Say what is needed and by when, give feedback quickly, and credit good work in public.

Communicate the decision, including a no. If your case is approved, say what happens next and who does what. If it is turned down, say what was decided, why, and what happens instead. Don't blame, and don't let the team find out sideways.

Look after the team. A launch adds work. Say what will stop to make room, and notice who is carrying too much.

AI activity. Ask an AI assistant to act as a sceptical finance director and challenge your case and risk register. Remove names and commercially sensitive figures first, or use public or invented ones. Decide which challenges are fair and answer them. Note any it raised that you had missed, and keep the challenges and your replies in your portfolio.

Milestone 4 is the last piece of your portfolio and the one you may actually use. It is one document that a decision-maker could read in the room. Read the worked example later in this topic first: it shows Maya's case for Larkfield.

The sections, in the order a reader needs them:

The executive summary. Write it last and keep it to one page. Open with the decision you are asking for and the date. Then the evidence, the recommendation with its ROI and break-even point, the main risk and what you need from the reader. If the reader stops there, they should still decide well.

Check the joins. The review's results feed the options. The costs match across the options, the financial case and the ROI. Every risk in the register has an owner who appears in the leadership plan. The recommendation is the option the economic case favours, or you say why not.

Ask for something specific. End with the decision, the date and the first three actions after a yes.