Which data source leads for revenue?

There is no single data source that leads for all revenue. The signed contract leads for what was agreed, the ERP or time tracking for what was delivered, the accounting system for what was invoiced and paid, and the CRM for what is still expected. So you choose the leading source per piece of data, not per system. If you declare one system the truth about all revenue, you fail to see exactly the differences where revenue leakage sits.

Why does this question keep coming back?

In almost every management meeting of a growing B2B company the moment comes: sales quotes one revenue figure, finance another. Then comes the question of which number is right. The answer is usually: both, for something different. Sales counts won deal value, finance counts invoiced revenue in the period. They are different questions that happen to both be called "revenue".

The reflex is to appoint one system. "The accounts lead, end of discussion." That makes meetings shorter. But it also means that anything that never reached the accounts no longer exists. A deal that was won but never invoiced then disappears not only from revenue, but also from the conversation.

Which source leads for which revenue question?

Revenue in fact answers four different questions. Each question has its own best source.

So the CRM does not lead for agreed revenue, even though the contract value is in it. The CRM is a copy, filled in by a salesperson, often before the last round of negotiation. The signed contract is the original. The article why CRM data is not the same as financial data explains why that copy differs so often.

The accounting system does not lead for what you should have invoiced. It leads for what you did invoice. The difference between the two is exactly the question.

The leakage sits between the sources

When you put the four sources side by side, three comparisons emerge. Each comparison catches a different kind of leak.

Agreed against invoiced. The contract says EUR 4,200 a month with annual indexation, the accounts invoice EUR 4,000 without indexation. This is where you find missed price indexation, discounts that keep running, wrong rates and contracts that are not (or no longer) being invoiced.

Delivered against invoiced. The ERP says 36 order lines were delivered, the accounts have invoiced 33. Time tracking says 180 billable hours, the invoice says 162. This is where you find forgotten invoices, uninvoiced extra work and missed final accounting on projects.

Expected against agreed. The CRM says the deal is won, there is no contract or order. Or the other way round: there is a contract, the CRM does not know about it. This is where you find deals stuck between sales and administration, and a forecast built on data that is wrong.

Each comparison requires that you can link the records, with a customer number, contract number or order number. The article CRM-to-billing reconciliation explained works through the third comparison step by step.

How do you record the choice?

An agreement that exists only in people's heads gets debated again at the first difference. Record it in a simple document, per piece of data.

1. List the data that determines revenue. Customer, contract value, unit price, discount, term, start date, end date, indexation, notice period, quantity delivered, hours worked, invoice amount, payment date.
2. Appoint one leading source for each piece of data. Not two. If the contract and the CRM both have a term, one of them is the original and the other a copy.
3. Appoint an owner for each piece of data. The person who updates the source and is responsible for its accuracy.
4. Record what happens when there is a difference. Who is alerted, within what time, and who decides whether the source or the copy is changed.
5. Record how copies are updated. If the contract leads for the term, how does a changed term reach the CRM and billing? Automatically, through a form, or through a task?

This document does not have to be long. A one-page table is enough. It determines more about the quality of your revenue figures than which system you use.

Exceptions and difficult cases

Verbal or emailed agreements. An account manager agrees a different rate by email. The email is then a contract change, but nobody treats it as one. The rule has to be: an agreement that changes the price or term is recorded in the leading source for agreements, or it does not exist. That also protects against the reverse situation: a customer invoking an agreement nobody can find.

Framework agreements without a fixed value. In a framework agreement the price is fixed but the volume is not. The contract then leads for price and the ERP for volume. The monthly expectation comes from history or from the CRM.

Usage-based billing. With licences or consumption, product data leads for usage, and the contract for the unit price and tiers. The invoice has to follow from those two.

Migrations. During a move to a new ERP or CRM there are temporarily two sources for the same data. Decide in advance the date on which the new source becomes leading, and compare both sources until every contract has been checked.

What does this mean for your reporting?

If you choose a leading source per piece of data, what you show in the management meeting changes too. Instead of one revenue figure that sparks debate, you show three figures side by side: agreed revenue for the period according to the contracts, delivered revenue according to the ERP and invoiced revenue according to the accounts. The interesting number is then not any of the three, but the difference between two consecutive ones.

A difference between agreed and delivered can be legitimate: a customer who postpones a delivery. A difference between delivered and invoiced rarely is. By showing the differences separately, the conversation shifts from "which figure is right" to "where does this difference come from and who fixes it". That is a better conversation, and it brings in money.

The same applies to the forecast. A forecast that relies only on the CRM carries the CRM's errors with it. A forecast that takes contractual revenue from the contracts and only new revenue from the CRM stands on firmer ground. That distinction comes back in what is revenue forecasting?.

Worked example

Worked example: suppose a B2B software supplier has 180 customers on annual contracts. For years the CRM was used as the leading source for contract value. When the signed contracts are compared with the CRM, it turns out that for 22 customers the contract value in the CRM is lower than in the contract, because the salesperson entered the value before the last expansion. Billing followed the CRM.

If the average difference is EUR 1,900 a year, that is 22 times EUR 1,900, so EUR 41,800 a year. As long as the CRM led, this difference did not exist: the accounts matched the CRM and everyone was satisfied. Only when the contract was appointed as leading did it become visible.

Frequently asked questions

Don't the accounts always lead?

For what was invoiced and paid, yes. For what should have been invoiced, no. If you use the accounts as your only source, you cannot see revenue leakage by definition, because the missing money was never in the accounts.

Can the CRM lead for contracts?

Only if the signed contract is recorded in the CRM and the fields are checked after signature. In practice the CRM is usually a copy filled in when the deal closed, and then the contract itself leads.

What if two sources are both partly right?

Then you have not split the data finely enough. Price and volume, or term and notice period, are different pieces of data with possibly different sources. Keep splitting until each piece of data has one source.

How does this fit into a wider approach?

Choosing leading sources is the basis of a single source of truth for revenue. That single truth is not one system, but an agreement on which source says what, plus a comparison that shows the differences.