“Free Business Case Template With Real Examples”

What sections does a business case template need?

The template needs eleven sections, in this order: purpose, problem, strategic fit, options, costs, benefits, funding, risks, delivery, recommendation, and assurance. Each answers one question a decision-maker will ask, and each should fail visibly if the evidence is missing.

The order matters. If you write the recommendation first, the rest of the document becomes advocacy. Write section 2 before section 10 and the case often changes shape.

The Five Case Model used across UK public sector splits the same content differently — strategic, economic, commercial, financial, and management cases — and it is worth borrowing its discipline even in a private company. The economic case forces you to compare options on value, not preference. The management case forces you to name who delivers it. Those two are where private-sector cases are usually thinnest.

How do you write the problem statement without adjectives?

Write the problem as a measured gap between where you are and where you need to be, with a date and a source. “Our Pickering warehouse runs at 94% capacity in November and December, up from 78% in 2022, and we turned away an estimated 1,900 orders last peak” is a problem statement. “Our warehouse is at breaking point” is a complaint.

Three tests separate the two. First, could someone else reproduce the number from your source system? Second, is there a trend, not just a snapshot? Third, does the gap have a cost — overtime, lost sales, penalties, turnover, safety incidents?

That third test produces the number reviewers care about most: the cost of doing nothing. If the status quo costs €380,000 a year in overtime and €210,000 in lost contribution, the project is not competing against zero. It is competing against a €590,000 annual bill that grows with volume.

Write that number down, date it, and put it in the summary. Everything else in the document is a comparison against it.

How many options should a business case compare?

Compare at least three: do nothing, a low-cost operational fix, and the capital option you actually favour. Two options invite a binary argument; four or more turn the document into a menu nobody reads.

The do-minimum option is the most useful and the most often faked. It should be a genuine attempt to solve the problem with process changes, better scheduling, or a small spend. If the low-cost option is credible, it raises the bar for the capital case, which is exactly what a reviewer wants. If it is not credible, say why in one sentence.

Then describe each option on the same five dimensions: capital cost, annual running cost, capacity or service level delivered, time to benefit, and reversibility. Reversibility is underrated. An option that can be undone in six months is worth more than one that locks the business into a building for fifteen years, even at the same NPV.

How do you value benefits you cannot invoice?

Value them through a proxy that finance will accept, then show the proxy. Four proxies cover most cases.

Avoided cost. Overtime, agency staff, penalties, subcontracting, expedited freight. Use last year’s actuals and adjust for volume. This is the most defensible category.

Time released. Hours × loaded hourly rate × weeks. If 8 people each save 3 hours a week for 46 working weeks at a €38 loaded rate, the annual value is 8 × 3 × 46 × €38 = €41,952. Do not claim this as cash unless someone actually removes the cost; otherwise label it “capacity released”.

Contribution margin. Extra units sold × margin per unit, not revenue. A project that adds €1m of revenue at a 22% margin is worth €220,000, and only if the demand exists.

Risk reduction. Expected loss avoided = probability × impact, using your own incident history. It is legitimate, but cap it at the value of the loss you can evidence.

Two rules keep this honest. Never count the same benefit twice — a headcount reduction and the time saving that justified it are one benefit, not two. And give every benefit a named owner and a measurement date, because a benefit without an owner is a hope.

Worked example one: a warehouse automation project

A regional distributor is at capacity. The proposal: install a shuttle racking system and pick-to-light in the Pickering site. Capital cost €1.8m, including racking, conveyors, software integration, and 14 weeks of disruption. The finance team’s hurdle rate is 8%.

The benefits, built from last year’s actuals: €380,000 of avoided peak overtime, €210,000 of contribution from orders currently turned away, and €90,000 of avoided third-party storage. Total annual net benefit: €680,000. Then the honest deduction: the system adds €160,000 a year in maintenance, software support, and extra energy. Net annual benefit: €520,000.

Sum of the discounted benefits: €2.076m. Net present value: €2.076m − €1.8m = €276,000. Payback: €1.8m ÷ €520k = 3.46 years. Benefit-cost ratio: 1.15. Internal rate of return: about 13.7%.

Now the part most cases leave out. If benefits come in 15% below plan — €442,000 a year — the present value falls to €1.765m and the NPV becomes −€35,000. The project fails. If capital cost overruns by 10% to €1.98m, the NPV falls to €96,000 but survives. This project is sensitive to benefit delivery, not to construction cost, and that should shape the approval conditions: phase the spend, pilot one aisle, and hold the second tranche until the pick rate hits a measured threshold.

That is the value of the exercise. The headline number says yes. The sensitivity table says “yes, if you can prove the pick rate in a pilot”. A reviewer can act on that.

Worked example two: replacing a boiler plant

A manufacturing site has a 22-year-old gas boiler plant. The proposal: replace it with industrial heat pumps and a heat-recovery loop. Capital cost €480,000, including enabling works and a 6-week shutdown managed around production.

Annual benefits: €95,000 of energy savings at current tariffs, €18,000 of maintenance and compliance savings. Total €113,000 a year. Asset life 20 years. The company uses a 6% real discount rate for energy projects.

The annuity factor for 20 years at 6% is 11.47. Present value of benefits: €113,000 × 11.47 = €1.296m. NPV: €1.296m − €480,000 = €816,000. Payback: 4.25 years. Benefit-cost ratio: 2.70. IRR: about 23%.

The undiscounted 20-year saving is €2.26m. The discounted figure is €1.296m, and the €964,000 difference is the cost of waiting. This is the number to show anyone who argues that discounting is an accounting technicality.

At a 10% discount rate the NPV is still €482,000, so the case survives a much harsher hurdle. That robustness — not the headline 2.70 ratio — is what makes it fundable. A case that only works at 6% is a case that depends on the discount rate, and discount rates move.

How do you present the numbers so finance can approve them?

Present four measures, one table, and one paragraph of assumptions. Finance does not need a model; it needs to see that you know which assumption kills the project.

NPV answers “does this create value at our cost of capital”. Payback answers “how long is our money at risk”. IRR answers “how much headroom do we have before the hurdle rate”. Benefit-cost ratio answers “how much value per euro”. No single measure is sufficient: a project can have a 40% IRR and a four-year payback on a small base, or a large NPV with a payback beyond the strategy horizon.

State the discount rate and where it came from. For public projects in the UK, HM Treasury’s Green Book applies a social time preference rate of 3.0% real for the first 30 years following the 2026 update, down from 3.5%. For a company, use the weighted average cost of capital or the rate your finance team already applies to capital requests. Do not invent a rate to make the NPV positive; that is the fastest way to lose the room.

Then include a sensitivity table with three variables: benefits −15%, capital +10%, and a one-year delay. Those three cover most of the risk in most projects, and the delay case is the one teams forget. A one-year delay on the warehouse project shifts every cash flow by a year: the NPV falls to roughly €163,000, because the €520,000 benefit loses a year of discounting and the spend does not.

What makes a business case fail in review?

Cases fail for predictable reasons, and almost none of them are “the idea was bad”.

The do-nothing option is a straw man. Reviewers notice when the alternative is described as “continued decline” with no attempt at mitigation. Build the best version of do-nothing, then beat it.

Benefits are counted at full value with no attribution. If a new system saves eight hours a week but the team already had slack, the benefit is capacity, not cash. Say which one it is.

Costs stop at go-live. The template should force a five-year total cost of ownership: licences, maintenance, energy, staff, training, decommissioning, and exit. A project with a €300,000 capex and €120,000 annual running cost is a €900,000 decision over five years, not a €300,000 one.

Nobody owns the benefit. A benefit register with names and dates is the difference between an approved case and an approved case that delivers. Assign each benefit to a role, not a department.

The recommendation is conditional on nothing. “Approve €1.8m” is weaker than “approve €600,000 for design and the first aisle, with the balance released at a measured pick rate of 180 units per hour”. Staged approval is easier to say yes to and easier to stop.

The case is never revisited. A business case is a living document in every serious method. Set gate reviews at design freeze, mid-delivery, and 12 months after go-live, and put the dates in the approval paper.

How do you handle benefits that cannot be monetised?

Put them in a separate, clearly labelled section and score them against the same criteria for every option, rather than inventing a cash value. Reviewers accept unmonetised benefits when the method is transparent and reject them when they are smuggled into the financial case.

A simple weighted scorecard handles this. Choose five criteria — say service quality, safety, environmental impact, workforce impact, and strategic alignment — weight them to sum to 1, and score each option 1 to 5. If the capital option wins on the financial measures and ties on the scorecard, the decision is easy. If the low-cost option wins the scorecard but loses on NPV, the document has surfaced a genuine trade-off instead of hiding it.

Two disciplines make the scorecard credible. Set the weights before scoring, and record who agreed them. Publish the unweighted scores alongside the weighted total so a reader can see whether the weights are doing the work. A scorecard that always produces the answer the author wanted is worse than no scorecard, because it looks objective.

Public-sector cases go further, with distributional analysis: who receives the benefit, who bears the cost, and whether any group is disproportionately affected. Even in a private company, one line on “who inside the business absorbs the disruption” improves the decision, because the answer is usually the team that was not consulted.

What does the one-page version look like?

The one-page version keeps the logic and drops the narrative: problem and baseline cost in two sentences, three options with capital and annual cost, the recommended option, NPV and payback, the top three risks, the benefit owner, and the decision requested with a date.

It works because it answers the same questions in the same order. What changes is the depth of evidence, not the structure. Use it for reversible decisions under roughly €100,000, or as the executive summary stapled to the front of the full case. Do not use it to skip the sensitivity analysis; one extra row is enough.

The test of a good one-pager is whether a director can read it in four minutes and ask one sharp question. If the question is “where did this number come from”, the appendix has failed. If the question is “what happens if the benefit is half of that”, the case is working.

How do you keep the case alive after approval?

Keep it alive by turning the benefits register into a project artefact with owners, measures, and review dates, then reporting against it at the same cadence as cost and schedule.

Three practices make that real. First, put the benefit measures in the project plan as milestones with a named measurer. Second, review the case at each gate and record any change in a decision log — scope changes without a case update are how projects drift from their justification. Third, do a post-implementation review at 6 and 12 months, comparing actuals to the case, including the ones that missed.

That review is uncomfortable and valuable. A team that publishes “we projected €520,000 and realised €410,000, here is why” earns the right to be believed next time. A team that never mentions it teaches the organisation that business cases are paperwork.

How does Doitify fit into this process?

Doitify is our product, so treat this as disclosure. It does not write a business case, and it will not choose your discount rate. What it does is hold the delivery side of an approved case: projects, tasks, sub-tasks, checklists, owners, due dates, Kanban and Gantt views, calendars, workload, project and performance reports, automations, documents, and Doitify Copilot for turning a goal into a plan.

The connection is the benefits register. If the case commits to a pick rate, a cost reduction, or a service level, those become measurable milestones with owners in the plan, reviewed at the same meetings as schedule and budget. Otherwise the case is filed, the project runs, and nobody compares the two until the post-implementation review — by which point the people who wrote it have moved on.

Conclusion

A business case template earns its keep by making the weak parts visible: an unquantified problem, a fake do-nothing option, a benefit with no owner, a cost that stops at go-live. Fix those four and the document becomes a decision tool rather than a persuasion exercise.

Use the eleven sections, write the problem before the recommendation, put the do-nothing cost in the summary, and include a sensitivity table with three rows. Then recommend the smallest commitment that produces evidence — a pilot, a first phase, a design stage — with the balance released on a measured result. That is easier to approve, easier to stop, and far more likely to deliver the number the case promised.

If this post on business case template was helpful, you might also enjoy How to Write a Project Status Report (5 Steps) and Free Work Breakdown Structure Template (Excel + Guide).