A CRM roadmap finance will fund

The CRM lead asks for two headcount and a budget line for tooling. The deck shows engagement trending up, a list of programmes shipped, and a revenue figure from the platform's own attribution. Finance asks one question: what happens if we do not approve this?

There is rarely a good answer to that question, and the request gets deferred to the next cycle. Not because the work is weak, but because the case was written in a language that does not survive the room it is presented in.

Finance is not hostile to retention spending. It is allergic to a number it cannot reconcile with the ledger. A CRM roadmap gets funded when it is priced the way every other investment in the business is priced, and when the person presenting it has already conceded the parts that cannot be proven.

Learn what the question actually is

"What is the ROI" is shorthand for four separate questions, and answering the wrong one is why these meetings go badly.

1. How much cash goes out, and when? Split by people, licences, agency and one-off build.
2. How much gross profit comes back, over what period, and how confident are you?
3. How much of that would have happened without the spend?
4. What is the alternative use of the same money, and why is this better?

Question three is the one CRM teams usually skip, and it is the one finance cares about most. Attribution answers a different question: which touchpoint was nearest the order. Nobody in a finance function believes that a discount code sent to a customer who buys weekly generated the whole order value.

Answer question three first, even if the answer is partial, and the rest of the conversation changes tone.

Price the plan in contribution, not attributed revenue

Revenue is the wrong unit. A retention programme that drives a million in sales at a heavily discounted margin may be worse than one driving half that at full price, and finance already knows this.

Convert everything to contribution before it goes in the deck:

• Start from gross profit, after cost of goods, not from order value.
• Subtract the discount or reward cost of the mechanic itself.
• Subtract fulfilment and payment costs where they vary with volume.
• Subtract returns at the rate the relevant cohort actually returns, which is often higher for discount-led purchases.
• Subtract the cost to serve, if the programme changes contact-centre load.

That arithmetic is where a lot of loyalty economics falls apart, as we worked through in the note on loyalty and margin. The same discipline applies to any lifecycle programme built on an offer.

When you express customer lifetime value in the plan, express it as contribution over a stated horizon, not as lifetime revenue. Twelve or twenty-four months is usually right. Anything longer is a forecast, and a forecast presented as a fact is what costs CRM teams credibility in this meeting.

Build the case on incrementality you have actually measured

The strongest slide in a CRM budget case is a small, honest incrementality result from work you already did. One programme, one holdout, one number, one caveat.

If you have it, lead with it. If you do not, the answer is not to estimate one from benchmarks. It is to say plainly that the current measurement cannot separate incremental from coincidental, and that part of this budget buys the ability to answer that question properly.

Finance responds well to that, in our experience, because it is the same argument they make about every other unproven line item. It converts the conversation from "trust our attribution" to "fund a measurement capability and hold us to it".

Where you have no experimental evidence yet, the next best support is a shape rather than a number. Show the retention curve for recent cohorts against older ones and point at the difference in the first ninety days. Show what a one-point improvement in ninety-day repeat rate is worth in contribution, using the arithmetic above.

Present that as sensitivity, not as a promise. "If the ninety-day repeat rate moves from its current level by one point, the annual contribution effect is roughly this, and here is the calculation" is a sentence finance can work with.

Write the roadmap as three tiers, not one number

A single ask invites a yes or no, and under pressure the answer is no. Three tiers invite a choice, and a choice is a decision somebody can own.

• Tier one, maintain. What it costs to keep the current programme running safely: licences, deliverability maintenance, the people who send. Describe what breaks without it, concretely.
• Tier two, fix the constraint. The one thing holding the programme back, usually data or measurement. One initiative, one owner, one outcome, one date.
• Tier three, grow. The new programmes, sized in contribution with explicit assumptions and an explicit measurement plan.

For each tier, name the quarter the work lands and the quarter the effect becomes measurable. Those are rarely the same quarter, and saying so in advance prevents the review meeting where a programme launched in March is judged on April's numbers.

Sequence tiers so the measurement capability in tier two lands before the growth spend in tier three. Funding growth before you can measure it is how a programme ends up back in this meeting next year with the same slide.

Agree the measurement before the money moves

The part that decides whether you are funded again is settled now, not at the review. Write it down with finance in the room and get it signed off alongside the budget.

Settle four things:

• The metric of record for each initiative, and who produces it. One source, one owner, not a platform report and a finance report that disagree.
• The comparison. A holdout where one is possible, a matched cohort where it is not, and a plain statement of which you are using and why.
• The window. When the result will be read, and the rule for what happens if the result is inconclusive.
• The kill criteria. What level of performance means the programme stops, decided before anybody is emotionally invested.

Then report against exactly that, monthly, in the same format every time, including the months that look bad. A CRM team that reports a miss before being asked about it gets the benefit of the doubt on the next ask.

Use cohort analysis as the standing view rather than month-on-month totals. Totals move with acquisition volume and mislead everyone; cohorts show whether the programme is changing customer behaviour.

If your funding cycle turns over in January, the planning work we set out for the January retention plan should be running in parallel with this. The budget case and the operating plan are the same document written for two audiences.

What to do this week

• Rewrite your top three initiatives in contribution terms, showing the arithmetic from gross profit down, and see which survives.
• Find the one piece of experimental evidence you already have, and if there is none, put a measurement line in tier two and say so out loud.
• Draft the three tiers on one page each, with the quarter the work lands and the quarter the effect can be read, and take it to finance before the formal submission.