How to Build a Business Case That Survives the CFO

The CFO Is a Quality Filter, Not an Obstacle

Reframe the audience before building anything. A CFO's job in your meeting is to find the case's load-bearing dishonesty before the organization funds it, and the honest cases move fast precisely because the dishonest ones are so common. The patterns finance teams see weekly: savings baselines built on the worst month of the year, benefits double-counted across cases, integration costs discovered post-approval, and point estimates that survived no one's audit. Your case competes against that memory.

The discipline that follows is almost negative space: one decision per case, ranges instead of points, downside modeled in the main body rather than a footnote, and every benefit carrying a method a stranger could audit. Get those right and the meeting changes from prosecution to confirmation, because you have removed the three things finance professionals are professionally required to attack.

One Case, One Decision

The first structural rule: a business case is a decision instrument, and it should argue exactly one decision. Approve or reject the warehouse automation at ninety thousand a year against option B and option C. Cases that bundle four decisions produce a single blended approval in which nothing is individually examined, and the weakest argument hides inside the strongest number. If the meeting needs to decide four things, bring four cases or one framework, but not a blend.

Within the single decision, present a real option set: do nothing (honestly costed, because the baseline is a choice too), the minimum viable version, and the recommended version with its variants. The do-nothing column is where credibility is won, because it shows the cost of inertia instead of pretending the status quo is free.

• One decision per case; split bundles ruthlessly
• Always include do-nothing as a costed option
• Minimum viable version before the recommended one
• Every option carries its own downside column

Baselines and Ranges: Where Cases Live or Die

The baseline deserves more hours than the forecast. Define it precisely (which twelve months, which cost centers, which allocation method), document it so a stranger could audit it, and include it as an appendix exhibit. Most case post-mortems trace back to a baseline chosen for attractiveness rather than accuracy, and one audited-baseline scandal poisons the well for your next three cases.

Then replace every point estimate with a range and a method: low, expected, high, with the assumption behind each. Point estimates invite binary approval; ranges invite a conversation about which world you are in, which is a better conversation to have in the room. State the swing factors explicitly: the two assumptions that most move the outcome deserve their own sensitivity table, and the assumption testing habit from strategy work applies here unchanged.

Model the Downside in the Main Body

The case that survives the CFO is the one where the CFO's question is already answered in the document: what happens if you are wrong? Give the downside its own section in the main body. If the expected benefit range fails to materialize, what is the break-even point, what is the exit cost, and what is the salvageable value? Present a stop-loss trigger if the decision warrants one, and name who holds it. Paradoxically, visible downside modeling raises approval odds, because it demonstrates that the case was built by someone who cannot afford to be naive.

The same honesty applies to time. Benefits that arrive in month eighteen should be shown arriving in month eighteen, with the ramp visible, because every finance team has watched the hockey stick that flattens after the approval meeting. If ramp assumptions are soft, say so and attach a milestone-based release of funds, a structure finance teams rarely refuse.

Pre-Wire the Room

The approval meeting is theater if the case is new to the room. The professional sequence: share the two-page draft with the finance partner a week early, explicitly invite the baseline audit, take version two into the influencers (operations, legal, the team that must live with it), and only then schedule the decision. By the meeting, the objectors have been converted into co-authors, and the case arrives with defenders already in the room.

Keep the artifact trail consistent with your wider strategy practice: the case's assumptions and stop-loss triggers belong in the same shelf as the trade-off logs and scenario triggers, so the organization builds one strategic memory rather than five. And when the case is really a strategy question wearing a spreadsheet, escalate it to the strategy document method instead of forcing it into an ROI box.

A One-Week Build Schedule

Monday: define the single decision and the option set, cost the do-nothing honestly. Tuesday: build the audited baseline and document the method. Wednesday: ranges and sensitivity table, downside section with break-even and exit costs. Thursday: pre-wire with finance, absorb the audit, revise. Friday: the two-pager is final, appendix complete, meeting scheduled. That cadence produces a case that survives contact with the CFO and, more importantly, produces decisions the organization can stand behind a year later.

If business cases and strategy artifacts are a recurring part of your role and the drafting load is the bottleneck, the AI-assisted workflow in Strategic Thinking & Business Strategy Mastery generates first drafts of every section, baselines, sensitivities, downside modeling, for your real numbers in minutes, leaving the judgment to you. Day 1 is free, and the scenario planning companion guide covers the uncertainty work that feeds harder cases.

Frequently asked questions