Executive summary
Build the case in a fixed order: problem and objective, option comparison including do-nothing, cost estimate with a stated accuracy class, benefit quantification tied to the baseline, risk register with mitigations, and financing route. Every figure should trace to a source, and the recommended option should be defensible against the strongest alternative, not only against inaction.
The situation this guide addresses
- Reviewers repeatedly ask where a figure came from
- Only one option has been costed in detail
- Benefits are stated without a baseline to compare against
- Risk section is a list of concerns without owners or mitigations
- Cost estimate accuracy class is not stated
- Financing route is assumed rather than justified
The framework, step by step
Step 1 — State problem and objective
One paragraph, measurable, tied to the baseline data already gathered.
Step 2 — Compare options including do-nothing
Cost, benefit and risk for each, on a consistent basis.
Step 3 — Estimate cost with a stated accuracy class
Declare whether the estimate is order-of-magnitude, budget or definitive.
Step 4 — Quantify benefits against the baseline
Tie every benefit line to the specific baseline metric it improves.
Step 5 — Build the risk register
List material risks with owner, likelihood, impact and mitigation.
Step 6 — State the financing route and sensitivity
Show the return under at least one adverse financing or demand scenario.
Comparison table
| Component | Minimum evidence standard |
|---|---|
| Problem statement | Tied to baseline data, one paragraph |
| Option comparison | At least two options plus do-nothing, consistent basis |
| Cost estimate | Stated accuracy class (order-of-magnitude, budget, definitive) |
| Benefit quantification | Traced to specific baseline metric |
| Risk register | Owner, likelihood, impact, mitigation per risk |
| Financing and sensitivity | At least one adverse scenario modelled |
Business case components and evidence standard
Risks and governance considerations
- A business case without a stated cost accuracy class invites unlimited later challenge
- Do-nothing should be costed, not dismissed, since it is the implicit alternative
- Benefits detached from the baseline are the most common reason cases are sent back
- A risk register without named owners is treated as incomplete by most approval bodies
- Sensitivity analysis on one variable at a time understates combined downside risk
What to prepare
- Baseline data and problem statement
- Costed option set including do-nothing
- Stated cost estimate accuracy class
- Risk register with named owners
- Sensitivity analysis on demand and cost variables
What to measure
Frequently asked questions
How detailed should the cost estimate be at business case stage?
A budget-level estimate, typically within plus or minus 15 to 25 percent, is normal before detailed engineering.
Is do-nothing really an option?
It should always be costed explicitly, since continuing constraints, downtime or lost margin are the true comparison baseline.
Who should write the risk register?
The project owner, in consultation with engineering, finance and operations, not a single function alone.
Related investment and financing knowledge
Project Lifecycle Stage
Business Case Development
Project Lifecycle Stage
Feasibility Study
Project Financing Knowledge
Financial Planning for Multi-Year Capital Programmes
Executive Decision Guide
How to Estimate Project Risk
CAPEX Intelligence
Industrial Investment Strategy
Project Lifecycle Stage
Project Concept
Executive Decision Guide
How to Prepare an Industrial Investment
Executive Briefing
What CEOs Own in Industrial CAPEX
Continue on the platform
Educational, supplier-neutral and financing-neutral
Global B2B Group does not sell equipment and does not represent lenders, export credit agencies or development banks. This material is published to help industrial organisations plan, structure and prepare capital projects. It is general information for decision-making, not financial, legal or tax advice.