A CRM year in review that finance will read

Every December the CRM team builds the same deck. Slide one has total attributed revenue and a large percentage next to it. Slides two to nine have campaign highlights, creative refreshes and a heat map of send times. Slide ten asks for more budget.

The finance director reads slide one, asks how attributed revenue relates to anything in the ledger, and does not get an answer she can use. The budget conversation then happens without the deck, on the basis of last year's number plus or minus a bit.

This is not because finance undervalues lifecycle marketing. It is because the deck is written in a currency finance does not accept. A year in review that changes a budget decision has to be written in theirs, and that changes what you put on nearly every page.

Attributed revenue is not a claim finance can check

Start by understanding the objection properly, because it is a good one.

When your platform reports that email drove a sum of revenue, it is applying an attribution model: a rule that assigns credit to touchpoints within a window. Change the window from seven days to thirty and the number moves. Change from last-touch to a positional rule and it moves again. Nothing about the business changed.

Worse, the number counts purchases that would have happened anyway. A customer who had already decided to reorder, who receives a reminder, clicks it and buys is recorded as email revenue. Some of that is real influence. Much of it is not, and nobody in the room can say which part.

Finance deals in numbers that survive a question about method. So lead with one that does.

Lead with incremental contribution from a holdout

The strongest number you can put on page one is the difference in revenue per customer between people who received the programme and a matched holdout group who did not.

If you ran a permanent global holdout this year, this section writes itself. State the holdout size, the period, revenue per customer in each group, and the difference with a confidence interval. Then multiply the per-customer difference by the mailed population to get annual incrementality in currency.

If you did not run one, do not manufacture a substitute. Say plainly that the programme's incremental value was not measured this year, report what you can defend, and put the holdout at the top of the plan for January. That admission costs you less credibility than a number that falls apart under one question.

Either way, keep the assumptions visible. Three lines are enough: what the holdout was, what period it covers, and what you multiplied by. An estimate with its method attached is arguable. A single large number with no method is dismissible.

Report cost the way finance reports cost

The CRM deck usually shows revenue and stops. Finance then has to find the cost themselves, and whatever they find will be less flattering than what you would have presented.

Put the full cost on the page yourself:

• Platform licence and overage, including the invoices that sit in someone else's cost centre.
• Deliverability, validation, SMS and push spend, separated by channel.
• Agency and contractor fees.
• Internal headcount at a loaded rate, including the design and development time the programme consumes from other teams.
• Incentive cost, meaning the actual discount and points value redeemed rather than the face value issued.

Then produce two ratios and explain both. Incremental revenue over total cost is the one finance will care about. Incremental gross margin over total cost is the one that is actually correct, and if you can get margin by product category from the finance team, use it. Asking them for that input also has a side effect worth having: they become a contributor to the analysis rather than an audience for it.

Show what you learned, not only what you earned

A year of tests is an asset, and most teams throw it away in January by never writing it down.

Include a test log as a plain list. Each entry needs the hypothesis, what was changed, the population, the result, and the decision that followed. Include the tests that failed, and especially the ones that failed expensively, because a team that reports its own negative results is a team whose positive results get believed.

Pull out the three findings that changed how the programme operates. Not "subject line B won by four per cent" but the structural lessons. A frequency ceiling you found. A segment that responds to service messages and not offers. A discount depth that made no difference to conversion.

This is also where you retire metrics that stopped meaning anything. We covered the practical version of that argument in life after open rate, and a year-end review is the natural moment to make the change formal rather than informal.

Reconcile with the numbers the business already uses

Your review will sit alongside a web analytics report and a finance report, and if the three disagree without explanation, yours is the one that gets discounted.

Do the reconciliation before the meeting. Take total orders in the period from the order system, and show how the analytics figure and your platform figure relate to it. Name the differences: attribution windows, bot filtering, cancellations and returns, test orders, channel definitions.

For most teams the analytics side of this changed this year, and if you rebuilt reporting during the migration described in reporting after Universal Analytics, say so explicitly. A year-on-year comparison across a measurement change needs a footnote, and providing it yourself is far better than having it discovered.

Set next year's targets in units you can be held to

End on commitments, because that is what the meeting is for.

Pick three or four targets and make each one measurable by someone outside the team. Incremental contribution per quarter, measured against a stated holdout. Twelve-month customer lifetime value for new cohorts, measured on a fixed window so cohorts stay comparable. Repeat rate within the first ninety days. Complaint rate ceiling by mailbox provider.

Then name what each target costs. A target with no resource attached is a wish, and finance can tell the difference immediately.

The hard part is not the deck. It is that a holdout you did not reserve in January cannot be recovered in December, and no amount of analysis at year end will produce evidence you did not set up. If you take one action from this piece before the break, make it the January holdout. Define the population, the exclusion rule and the review date, and get it agreed while budgets are still being written.