The CFO Business Case: How to Build an ROI Model a Finance Leader Will Actually Sign

One page, and a CFO will read it in four minutes. The problem, stated in money per month, not in adjectives. The cost of the status quo over the horizon you are proposing. The total cost of your solution over the same horizon, fully loaded. The expected return, as a range with a base case and a downside case. The payback period. The key assumptions, listed so they can be challenged. And the exit: what happens if it does not work, and how much has been lost by then.

Everything else, the product detail, the architecture, the customer logos, belongs in an appendix the CFO will not read and the champion will. A business case that mixes the two makes finance hunt for the numbers, and a CFO who has to hunt assumes they are being hidden.

Nothing in the case matters more than this line, because it converts your product from a new expense into the end of an existing one. Work with your champion to establish what the current approach costs each month: people, tools, vendors, the hours lost, the revenue not won, the errors and their cost. Be conservative and show the working. A CFO who agrees with your status-quo number has already accepted that there is a problem worth spending money on, and the rest of the conversation is about how much.

Where the current cost is hidden inside salaries, load it honestly. Finance leaders know what their people cost and will not accept a case that treats existing headcount as free, but they will accept one that says 'three people spend a third of their time on this, which is roughly $110,000 a year'.

List every dollar the company will spend to get the outcome, not just the invoice you send. Implementation, integration, training, the internal owner's time, any change to other tools, and the cost of switching away later. If the honest number is higher than your list price, that is fine: the CFO will find it anyway, and a case that found it first is a case they trust. A case that hides it loses credibility on every other number too.

Show the cost per period against the return per period, not just totals. Finance runs on monthly and quarterly views, and a purchase that looks large as a three-year total can look obviously right as a monthly line next to the status quo it replaces.

Return on investment is a marketing number; payback period is a finance number. A 300% three-year ROI with a thirty-month payback is a worse proposal to most CFOs than a 60% ROI that is cash-positive in four months, because the second one is cheap to be wrong about and the first is not. State the month in which cumulative cash saved or earned exceeds cumulative cash spent, and if that month is inside a year, put it in the first line of the case.

If you cannot get payback inside a year, restructure the offer until you can, by reducing the initial commitment, phasing the rollout, or moving to outcome-based pricing. Changing the shape of the deal is almost always more persuasive than defending the size of it.

A single forecast is a guess; a range with visible assumptions is an analysis. List the three or four inputs that drive the return, the adoption rate, the conversion lift, the hours saved, and show a base case and a downside case where each is cut by half. If the downside case still pays back, say so prominently; it is the sentence a CFO is looking for. If it does not, say what changes and what the exit costs.

Invite the CFO to stress the assumptions in the meeting. A finance leader arguing with your model is a finance leader deciding to buy; the ones who are not going to sign do not bother.

Every CFO knows that declining a purchase costs nothing and approving a bad one costs a lot, so the asymmetry favours no. Reverse it by making yes cheap to undo. Month-to-month terms rather than annual commitments, a capped budget the company controls, spend that only happens when an outcome does, and unused money that comes back. A reversible purchase gets approved on a much weaker business case than an irreversible one, because the downside line is small enough not to need a debate.

'It is not in the budget.' The answer is the status-quo line: the money is already being spent, just on a worse version of the outcome, and the case reallocates it. 'Show me who else has done this.' Bring one comparable company with a number, not a logo wall. 'What if it does not work?' Point to the downside case and the exit, and offer the reversible structure before they ask for it. 'Our team could do this in-house.' Agree, then load the in-house cost fully: salaries, ramp time, management, tooling, and the months of delay, and let the comparison sit next to yours.

'Why now?' Quantify the cost of each month of delay from the status-quo line; if waiting is free, they will wait. 'Is this capex or opex, and how does it hit the forecast?' Know the answer for your product and offer both shapes if you can; a finance leader who hears you understand their statement stops treating you as a vendor. 'What is the exit?' State the notice period, the data return and the switching cost in one sentence. If you have to think about it, you have already lost a point.

Take the business case a sales leader brings to their CFO for sourcing qualified meetings, because it is one we see constantly and because it shows the framework end to end. Status quo: an SDR at a fully loaded $80,000 to $120,000 a year, plus $500 or more a month in outbound tooling, plus management time and three to six months of ramp, producing a variable number of meetings. Divided by qualified meetings actually held, most companies land at $600 to $1,200 per meeting, with high variance and no way to scale down quickly.

Proposed: a pay-per-meeting marketplace. On Bountii the fully loaded cost is a flat plan of $199 a month for ten target accounts or $499 for thirty, plus a bounty the company sets per qualified meeting, typically $300 to $500 for mid-market decision-makers and $500 to $1,000 for enterprise executives. Payback is immediate in the sense finance cares about: the bounty is only charged after a meeting on a named account has happened, passed the criteria fixed at posting, and been reviewed within 48 hours. Assumptions are few and visible: contract value times meeting-to-close rate gives the ceiling price of a meeting, and the bounty sits well under it. Reversibility is built in: month-to-month plans, a bounty pool funded upfront so spend is capped by design, and unused balance refundable.

That case gets signed not because the numbers are dramatic but because every line answers a question the CFO would otherwise have to ask. It is also the reason we priced Bountii this way. Companies can post their first bounties from October 19, 2026, and the business case you take to your own CFO can be the one above.

A CFO signs the business case they could have written themselves. Price the status quo, load your full cost, lead with payback, expose the assumptions with a downside case, and make the purchase reversible. Then answer the seven objections before they are raised. Do that on one page and finance stops being where deals go to die and starts being where they get approved.