Hospital digital business case: what the CFO needs

A CFO once stopped me four slides into a digital investment case with a question I could not answer: “If I approve this, which line in the monthly pack changes, and by when?”

I had enquiry projections, a funnel, a competitive comparison and a view on market share. None of it mapped onto anything he looked at. He was not being obstructive. He was telling me, accurately, that my proposal was unreviewable — that having approved it, he would have no way of knowing whether it had worked, and therefore no way of defending it in a board meeting eight months later.

That conversation changed how I work more than any course. The obligation runs in one direction here. The finance function does not owe you fluency in digital. You owe them fluency in theirs.

What the CFO is actually solving for

Not cost minimisation. That is the caricature, and believing it will make you a poor partner.

A hospital CFO is managing four things simultaneously: the predictability of earnings, the shape of the cash flows, the defensibility of every number that leaves the building, and the capital allocation queue. Your proposal competes in that last one — and it does not compete against other marketing ideas. It competes against a cath lab, a new linear accelerator, a floor of beds at a unit running near capacity, and the working capital locked up in payer receivables.

That comparison is brutal and it is also winnable, because equipment has a long payback and a large capital outlay, and digital usually does not. But you only win it if you present in the same units as the competition.

Timing matters as much as content. A proposal that arrives while the annual plan is being assembled is a line item under discussion; the same proposal in the middle of the year is an unbudgeted request, and it will be judged far more harshly for the same money. Learn your group’s planning calendar — when unit budgets are submitted, when the capital list is frozen, when the board approves — and put your asks into it rather than into whatever month you happened to finish thinking.

The cost lines finance actually watches in a hospital

Learn these properly, in your own group’s chart of accounts, before you write a business case. Ask for the monthly management pack and read twelve months of it.

Occupancy and revenue per occupied bed. The two numbers that drive almost everything else. Any initiative that credibly touches either gets attention immediately.
Payer mix. Cash, insurance through third-party administrators, corporate, government scheme. These carry very different realisation and very different collection timelines. An initiative that brings volume but shifts mix towards slower-paying payers can be worse than no initiative, and finance will see that before you do.
Receivables and collection days. If your group has significant money sitting in TPA receivables, anything digital that reduces claim rejection or documentation delay is far more valuable than incremental enquiries, and it is almost never in a digital team’s plan.
Doctor cost. However it is structured — fixed, fee for service, revenue share — it is one of the largest lines and it behaves differently by unit and specialty. Volume in a high-payout specialty contributes less than the top line suggests.
Concessions and discount leakage. Finance watches this closely. A digital campaign promoting a package is, from their side, a structured discount, and they will want to know the contribution margin, not the package price.
Marketing and business development spend as a share of revenue. In hospital groups this typically sits in the low single digits, and the CFO knows the benchmark range for peers. Arguing for a number far outside it requires an explanation, not enthusiasm.

If you can discuss your own group’s position on these six without a slide, the relationship changes. You stop being a spender and start being someone who understands the business they are spending in.

Capex versus opex, and why it changes what you build

This is not an accounting footnote. In practice it determines your architecture.

Capital spend is approved in a different forum, on a different cycle, with a higher threshold and a longer memory. It is depreciated, which means it sits in the P&L for years and shows up as an asset on the balance sheet. Operating spend hits this year’s earnings directly and is reviewed every month, but it is easier to start and easier to stop.

The consequences are concrete.

A licensed platform with internal development effort may qualify, in part, for capitalisation. A subscription service generally does not. Two solutions that look equivalent to you have quite different effects on reported earnings, and the CFO may well have a preference.
Capitalised software that you abandon in year two creates a write-off conversation. That is a genuinely unpleasant meeting, and it is a reason to be conservative about large custom builds early.
Subscription-heavy stacks are pleasant to start and accumulate. Finance will eventually notice the aggregate and ask why it has grown faster than revenue. Keep your own inventory of every recurring digital subscription across every unit, with renewal dates. Be the one who brings that list before it is requested.

Ask your finance team where the line sits in your group’s policy before you scope a project, not after. I have had a proposal restructured twice, not because the business case was weak, but because the spend shape was wrong for that year.

What makes a business case defensible

Defensible is the key word. The CFO is not asking whether you believe the case. They are asking whether it survives a board member’s question and an auditor’s eye.

Four properties do the work.

A single named measure that changes in the monthly pack. Not a basket. One line, stated upfront, with a date. If you cannot name it, the case is not ready.

Attribution you have agreed in advance with someone other than yourself. Sit with the finance analyst who prepares the pack. Agree exactly how a digitally originated case will be identified in the hospital system, who runs the query, and what happens to cases that cannot be classified. Write it down and have them acknowledge it. A claimed result that finance cannot reproduce is worse than no result, because it costs you credibility on everything afterwards.

Contribution, not revenue. Revenue sourced through a channel is a number digital teams love and finance discounts heavily. Work with them to get to contribution per case for the specialties you are targeting, even roughly. A smaller honest number beats a larger one that gets adjusted in front of the board.

A stated downside and a stopping rule. What you expect if this underperforms, what the early signal would be, and at what point you would recommend stopping. Volunteering a stopping rule is the single most effective thing I have done in budget conversations. It signals that you are managing the group’s money rather than defending your function’s territory, and it makes approval a smaller decision.

What to leave out

Some things actively weaken a case in front of a finance audience.

Industry benchmarks as justification. “Peer groups spend more on digital” invites the response that peer groups also have different payer mixes and different capital positions.
Brand and awareness outcomes as primary benefits. They are real, they are long-term, and they are unfalsifiable in a quarterly review. Mention them as secondary, never as the case.
Cost savings you cannot point to in an account. Efficiency that does not reduce a line item is not a saving, it is a hope. If you are claiming lower cost per acquisition, say which budget line goes down or which volume goes up at the same spend.
Vendor-supplied projections. Using them tells the CFO you have outsourced your judgement. Build your own, and show the assumptions separately so they can be challenged individually.

When the payback is longer than the budget cycle

This is the real problem in hospital digital. Content and organic discovery take several quarters to compound. A contact centre redesign takes a year to show in conversion. Platform consolidation may show nothing in year one except avoided future cost. Meanwhile the budget cycle is annual and the review is monthly.

Four approaches that have worked, roughly in order of preference.

Split the case into a funded experiment and a conditional scale-up. Ask for a small, bounded amount with defined checkpoints, and pre-agree what evidence unlocks the next tranche. You are converting one large uncertain decision into a sequence of small ones, which is how capital committees prefer to work anyway.

Pair the long initiative with a short one. Fund the slow compounding work alongside something that moves a visible line within a quarter. This is not a trick; it is how you keep a long programme alive through the months when it has nothing to show.

Use leading indicators finance has agreed to accept. If a twelve-month outcome is the real measure, agree at the start which three-month signals count as on-track. The agreement must happen before the spend, never during the defence.

Reframe from growth to avoided cost or protected revenue. Some long-horizon digital work is genuinely defensive — protecting referral flow, reducing dependence on aggregator commissions, keeping enquiry capture when a competitor opens nearby. Defensive cases are easier to fund on a long horizon than growth cases, because the CFO already believes the threat.

What I would avoid: asking for a multi-year commitment upfront to protect a programme from annual scrutiny. It feels safer and it is not. You will get less money, with more conditions, and the first bad quarter will reopen the whole thing anyway.

What to send without being asked

Most of the relationship is built between budget cycles, in a monthly habit.

Actual spend against approved, by unit and by line, before finance has to chase it.
Anything you are stopping, and the money it frees. Returning budget buys more credibility than any result.
Your subscription and contract inventory with renewal dates, quarterly.
One page on what underperformed and what you are changing.
Early warning of anything that will need capital approval two cycles ahead, so it is never a surprise.

The objective is that your numbers never contradict theirs in a board meeting. That happens once before it becomes the thing you are known for.

If you’re going into the budget cycle next quarter

Get twelve months of the management pack and read it properly. Identify the three lines your work can plausibly touch.
Ask finance where the capitalisation line sits in group policy, and scope accordingly.
Sit with the analyst who prepares the pack and agree attribution mechanics in writing.
Build every case around one named measure, with a date and a stopping rule.
Convert at least one initiative into contribution terms, even approximately, with finance’s help.
Pair each long-horizon programme with a short-horizon one, and pre-agree the leading indicators.
Bring your subscription inventory and a list of what you propose to stop. Lead with that, not with the ask.

The digital leaders who get funded are not the ones with the best story. They are the ones whose last three numbers held up when somebody checked.

Questions people ask

What does a hospital CFO actually want from a digital investment proposal?

One line in the monthly management pack that will change, and a date by which it changes. A hospital CFO is managing earnings predictability, cash flow shape, the defensibility of every number and the capital queue. Your proposal competes in that queue against a cath lab or a floor of beds, not against other marketing ideas. It is winnable, because digital has a shorter payback than equipment — but only if you present in the same units as the competition.

Which hospital P&L lines should a digital head understand before writing a business case?

Six. Occupancy and revenue per occupied bed, payer mix, receivables and collection days, doctor cost, concessions and discount leakage, and marketing spend as a share of revenue. Ask for twelve months of the management pack and read it properly. If you can discuss your group’s position on those six without a slide, the CFO stops seeing you as a spender and starts seeing someone who understands the business they are spending in.

Does it matter whether hospital digital spend is capex or opex?

Yes, and in practice it decides your architecture. Capital spend is approved in a different forum, on a longer cycle, and sits in the P&L for years through depreciation. Opex hits this year’s earnings but is easier to start and stop. A licensed platform with internal build effort may partly capitalise; a subscription generally does not. Ask finance where the line sits in group policy before you scope, not after — I have had a case restructured twice for spend shape alone.

How do you fund a digital programme whose payback is longer than the annual budget cycle?

Four ways, roughly in order of preference. Split the case into a funded experiment and a conditional scale-up with pre-agreed checkpoints. Pair the slow compounding work with something that moves a visible line within a quarter. Agree leading indicators finance will accept before the spend starts. Or reframe from growth to avoided cost or protected revenue, which CFOs fund more readily on long horizons because they already believe the threat.

What makes a digital business case defensible in front of a hospital board?

Four properties. A single named measure in the monthly pack, with a date. Attribution mechanics agreed in advance with the finance analyst who prepares the pack, written down and acknowledged. Contribution per case rather than channel revenue, even roughly. And a stated downside with a stopping rule. Defensible means it survives a board member’s question and an auditor’s eye eight months later — not that you believe it.

What should a digital head leave out of a business case to the CFO?

Industry benchmarks as justification, because peers have different payer mixes and capital positions. Brand and awareness as the primary benefit, because they are unfalsifiable in a quarterly review. Cost savings you cannot point to in an account — efficiency that does not reduce a line item is a hope, not a saving. And vendor-supplied projections, which tell the CFO you have outsourced your judgement. Build your own and show the assumptions separately.

How do you agree attribution for digitally sourced patients with hospital finance?

Sit with the analyst who prepares the monthly pack, before the spend. Agree exactly how a digitally originated case is identified in the hospital system, who runs the query, and what happens to cases that cannot be classified. Write it down and have them acknowledge it. A claimed result that finance cannot reproduce is worse than no result, because it costs you credibility on everything that follows.

Why should a digital head volunteer a stopping rule in a budget conversation?

Because it makes approval a smaller decision. State what you expect if the programme underperforms, what the early signal would be, and at what point you would recommend stopping. It is the single most effective thing I have done in budget conversations. It signals you are managing the group’s money rather than defending your function’s territory, and a CFO who trusts that will fund the next case with far less friction.

What is contribution per case and why does finance prefer it to channel revenue?

Contribution is what remains of a case’s revenue after the direct costs of delivering it — doctor payout, consumables, concessions. Revenue sourced through a channel is the number digital teams love and finance discounts heavily, because volume in a high-payout specialty contributes far less than the top line suggests. Work with finance to get to contribution for the specialties you target, even approximately. A smaller honest number beats a larger one adjusted in front of the board.

When should a digital team submit budget requests in a hospital group?

Inside the planning calendar, not whenever you finish thinking. A proposal that arrives while the annual plan is being assembled is a line item under discussion. The same proposal mid-year is an unbudgeted request and is judged far more harshly for the same money. Learn when unit budgets are submitted, when the capital list freezes and when the board approves, and give finance two cycles’ warning of anything needing capital.

Is asking for a multi-year digital budget a good way to protect a long programme?

No. It feels safer and it is not. You will get less money, with more conditions, and the first bad quarter reopens the whole commitment anyway. Convert one large uncertain decision into a sequence of small funded ones with checkpoints instead. That is how capital committees prefer to work, and it keeps the programme alive through the months when it has nothing visible to show.

What should a digital head send the CFO every month without being asked?

Actual spend against approved, by unit and by line, before finance has to chase it. Anything you are stopping and the money it frees — returning budget buys more credibility than any result. A subscription and contract inventory with renewal dates, quarterly. One page on what underperformed and what you are changing. And early warning of any capital need two cycles ahead. The objective is that your numbers never contradict theirs in a board meeting.

Can a vendor’s ROI projection be used in a hospital digital business case?

Not as the case. Vendor projections are built to sell, and presenting them tells the CFO you have not formed your own view. A good vendor gives you the inputs — unit costs, realistic ramp times, reference ranges — and lets you build the model with your own assumptions shown separately, so each can be challenged individually. If a vendor cannot support that, treat it as a signal about how the relationship will go after signature.

Does this apply to a single hospital or a Tier 2 unit with a small digital budget?

More so, because the CFO’s attention is closer and the numbers are smaller. In a single unit, the six lines are easier to learn and the attribution conversation is with one analyst. Marketing spend as a share of revenue is watched just as closely. The discipline is the same: one named measure, agreed attribution, a stopping rule, and a subscription inventory brought before it is asked for. Small budgets are lost to sloppiness faster than large ones.