Revenue Intelligence for SaaS

Revenue Intelligence for SaaS means continuously comparing three sources that rarely agree at a software company: what sales sold in the CRM, what billing actually invoices and what customers use in the product. Where those three diverge, there is revenue you are missing or forecasting wrongly. A well-designed approach tells you not only how high your ARR is, but also which part of it is right, which part is at risk and which part you are not yet billing.

Why does SaaS need its own approach?

On paper, a SaaS company looks like the most measurable business model there is. Everything is digital, every customer has a subscription, every invoice comes out of a system. Yet it is exactly that sense of control that makes it dangerous. Because billing runs automatically, nobody checks any more whether it is automatically doing the right things.

Three characteristics set SaaS apart from a project or trading business.

Revenue is a stream, not a moment. In SaaS, a deal in the CRM is not a one-off gain but the start of a series of invoices that can run for years. An error at the start therefore repeats every month or every year until someone spots it. A wrongly entered 10 percent discount on a contract of EUR 30,000 a year does not cost you EUR 3,000, but EUR 3,000 a year for as long as the customer stays.

The value sits in three systems at once. The CRM (for example Salesforce or HubSpot) knows what was agreed. Billing (Stripe or Chargebee, or invoicing from an accounting package such as Xero, NetSuite, Exact or Moneybird) knows what is invoiced. The product knows what the customer uses: seats, API calls, storage, modules. None of these systems has the whole story, and none of them checks the other two.

Your valuation depends on metrics built on this data. In SaaS, ARR, net revenue retention, gross retention, churn and expansion are not just management information. They determine how investors, buyers and banks look at your business. If the ARR in your board deck comes from the CRM and the ARR in your accounts comes from billing, and the two differ, you have a problem bigger than a few missed invoices.

What Revenue Intelligence is in general is explained in the complete guide to Revenue Intelligence. This article covers what it means specifically for a SaaS company.

The three sources and where they contradict each other

CRM: what you think you sold

The CRM holds the deal with an amount, a term, a product line and often an ARR field someone fills in by hand. That field is the start of a lot of trouble. It is filled in at closed-won, but rarely updated at an upgrade, a downgrade or a mid-term change. After a year the CRM describes the customer as they were at signing, not as they are now.

Typical discrepancies between the CRM and reality in SaaS:

• A multi-year contract with a stepped price (year 1 EUR 24,000, year 2 EUR 30,000) is recorded as a single amount in the CRM.
• A first-year discount is in the notes, not in a field, so nobody knows when it ends.
• An upsell was created as a new deal, but the original contract was not updated, so the ARR is counted twice or not at all.

Billing: what you actually invoice

Billing is the source for what comes in. But billing does what was set up, not what was agreed. At some point someone created a subscription, chose a price, applied a coupon and set an invoice date. Every error in that step runs on automatically from that moment.

The classic example: sales sells 50 seats at EUR 40 per seat per month. In billing, the subscription is created on the standard plan with 25 seats, because the custom plan did not yet exist. Nobody notices, because the customer does not complain and the invoice goes out neatly every month. That is EUR 1,000 a month, EUR 12,000 a year, that never shows up anywhere as a loss.

Product usage: what the customer really consumes

The product is the third source and in SaaS often the most underrated. It tells you how many users are active, whether a customer is above their limit and which modules are used. This is where you find both leakage (usage above what you bill) and risk (usage that falls well before a cancellation).

A customer paying for 50 seats and using 72 is leaking revenue. A customer paying for 50 seats and using only 11 will probably scale down or leave at renewal. Both signals are in your product data. Neither is in your CRM.

Where does SaaS revenue leak?

Leaks in SaaS are rarely spectacular. They are small discrepancies that repeat every billing period. A fuller list by cause is in revenue leakage in SaaS. The main patterns:

1. Usage above contract. More seats, more volume or more API calls than billed. Often because the product has no hard limit and billing does not look at usage.
2. Discounts that never end. A 20 percent introductory discount for the first year that sits in billing as a permanent coupon.
3. Price increases not applied. The new price list applies to new customers, but existing contracts with an indexation clause stay on the old rate because nobody executed the clause.
4. Mid-term upgrades without pro rata billing. A customer on an annual plan moves to a higher tier in month four. The upgrade is activated, but the extra charge for the remaining eight months is never invoiced.
5. Failed payments. A credit card expires, the direct debit fails, the dunning emails land in a spam folder and the account stays active. This is called involuntary churn if the customer is eventually cut off, but it is leakage for as long as the account keeps running without payment.
6. Custom arrangements outside the system. An implementation fee, an extra module or a professional services block that was in the quote but was never created as an invoice line.
7. Renewals at the old rate. A contract renews automatically, but the renewal is set at the old price, without the indexation or the new price list.

These patterns largely fall under what is called SaaS billing leakage: the gap between what you should have invoiced and what billing did invoice.

The metrics you trust, and why they can mislead

SaaS companies steer on a small number of core figures. Revenue Intelligence does not change those figures, but it checks what they rest on.

The point of this table: if you take ARR from the CRM and NRR from billing, you are comparing two different realities. Revenue Intelligence forces the two into one definition by putting them side by side per customer. How to approach that technically and organisationally is covered in how do you get a single source of truth for revenue.

Reconciliation: the core of Revenue Intelligence in SaaS

The work that pays off most is dull: per customer, per month, comparing what was agreed, what was invoiced and what was used. That is called SaaS revenue reconciliation. In practice it consists of four comparisons.

Contract versus subscription. Does the subscription in billing match the contract? Unit price, number of units, term, billing frequency, discounts with an end date, indexation clause.

Subscription versus invoice. Did the subscription actually produce an invoice, for the right amount, in the right period? This is where you find failed payments, manual credit notes that were never reversed and invoices left in draft.

Usage versus subscription. Is the customer within what they pay for? This is where you find seats, volume and modules above contract, and the reverse: customers using far less than they pay for, which is a renewal risk.

Invoice versus payment. Has the invoice been paid? An unpaid invoice from an active account is not an accounting detail, it is a customer using the product for free.

Worked example: a SaaS company with EUR 6M ARR

Take 400 customers and an average contract value of EUR 15,000 a year. You run a reconciliation once a quarter and find the following. These are example figures to show the calculation, not averages.

• 18 customers use on average 12 seats more than they pay for, at EUR 35 per seat per month: 18 x 12 x EUR 35 x 12 = EUR 90,720 a year.
• 9 customers have a first-year discount of 15 percent that stayed in place after year one, on an average of EUR 20,000 a year: 9 x EUR 3,000 = EUR 27,000 a year.
• 30 contracts with a 3 percent indexation clause have not been indexed, at an average of EUR 12,000: 30 x EUR 360 = EUR 10,800 a year, and that amount compounds every year it is left.
• 6 accounts with a failed payment have been active for two months without paying, at an average of EUR 1,500 a month: EUR 18,000 in outstanding revenue.

Together that is over EUR 140,000, around 2.4 percent of ARR. That falls within the range of 1 to 5 percent of revenue commonly cited as an estimate for revenue leakage. For your business it may be more or less. The only reliable figure is the one you find in your own data.

Forecasting and renewals

In SaaS, the largest predictable revenue item is not new business but renewals. Yet the forecast is often built on new pipeline in the CRM, with renewals as a fixed assumption. That works until a few large customers scale down at the same time.

Revenue Intelligence improves the renewal forecast by bringing in usage data. A customer with falling usage, rising support tickets and a contract ending in four months is not a certain renewal, even if the CRM shows them as "renewal, 90%". Conversely, a customer structurally above their contract is an expansion opportunity that belongs with the renewal, not with a separate sales campaign.

Two principles:

• Build the renewal forecast per customer, not as a percentage. An assumption of 92 percent gross retention says nothing about which customers will drop off. A list of risk signals per customer does.
• Separate contracted revenue from expected revenue. A signed multi-year contract is something different from a monthly subscription that can stop tomorrow. Do not add them up as if they were equally certain.

A forecast that treats renewals as a fixed assumption misses exactly the customers who make the difference.

Churn and expansion: signals before it is too late

Most SaaS companies measure churn after the fact: who cancelled this quarter? By then the revenue is gone. Revenue Intelligence looks at the signals that come before it, across systems.

Signals of risk:

• Active users fall three months in a row.
• The main user or the decision-maker has left (new email address, old login inactive).
• Support tickets rise, or response times for this account are above your standard.
• Invoices are paid later than before.
• The customer asks for a copy of their data or for the export function.

Signals of expansion:

• Usage is structurally above the contract.
• New departments or sites are logging in.
• The customer uses a feature that belongs to a higher tier, through a trial or a workaround.

Each of these signals can be seen on its own in one system. The value lies in the combination, and in the question of who acts on it. A signal without an owner is a report. A signal with an owner and a deadline is an action.

How do you get started?

You do not have to start with software. The first round can be done with an export and a spreadsheet, as long as you know what you are looking for.

1. Choose one definition of ARR. Set out what counts (recurring subscription revenue, after discounts) and what does not (one-off fees, professional services, uncontracted usage). Write it down.
2. Build a customer list with one key. Make sure every customer in the CRM, billing and the product can be matched with the same ID. If that is not possible, this is your first project.
3. Export three amounts per customer. Contract value from the CRM, invoiced amount from billing over the last twelve months, and usage (seats or volume) from the product.
4. Flag every discrepancy above a threshold. For example, more than 5 percent difference between contract and invoice, or usage more than 10 percent above contract.
5. Check discounts and indexations separately. List all active coupons and discounts with their original end date, and all contracts with an indexation clause.
6. Assign an owner to each discrepancy. Finance for invoice errors, customer success for usage above contract, sales for contract changes.
7. Repeat it every month. Checking once finds the old leaks. Checking monthly prevents new ones.

If you want this worked through thoroughly once, you can opt for a standalone audit. The Revenue Audit is a one-off review, carried out personally, across eight areas, in which every finding is given an amount in euros. If you want to monitor it continuously, you need software that connects the systems and compares them every night. Those are two different choices, and you can make one without the other.

What does a platform add over a spreadsheet?

A spreadsheet works for a first round. The problem starts with the second. The export is a snapshot, the matching between systems is done by hand, and after three months the person who did it is busy with something else.

A Revenue Intelligence platform for SaaS essentially does four things:

• It reads the systems continuously, instead of an export each quarter. A discrepancy from last week shows up this week, not at year-end close.
• It links at customer level, so contract, subscription, invoice and usage for the same customer sit side by side.
• It prices the discrepancy, so you know which ten of the hundred signals are worth the money.
• It turns it into a task with an owner, instead of a report nobody reads.

What it does not do: replace your billing, replace your CRM or decide your definitions for you. The systems you have remain the source. The platform puts them side by side.

Common mistakes at SaaS companies

Reporting ARR from the CRM. The CRM knows what was sold, not what is running. Use billing as the source for running ARR and the CRM for new and expected revenue.

Usage-based pricing without checking the metering. If you bill on usage, the measurement in the product is your invoice. A bug in the count or a missed export is a direct loss of revenue. Check periodically that the measured volume and the billed volume are equal.

Not recording discounts with an end date. Every discount has a reason and a duration. A discount without an end date in the system becomes a permanent price cut.

Not putting custom deals back into the system. The largest customers often get bespoke terms. That is exactly where the chance of a gap between contract and billing is highest, and where the amount is highest.

Looking at churn only per quarter. Then you only see who has gone. Look monthly at the signals of who is going.

Frequently asked questions

What is the difference between Revenue Intelligence and a SaaS metrics dashboard?

A metrics dashboard shows ARR, churn and NRR based on one source, usually billing. Revenue Intelligence puts the sources side by side and shows where they differ, per customer and in euros. The dashboard tells you the number, Revenue Intelligence tells you whether it is right.

Which systems should a SaaS company connect?

At a minimum the CRM, billing and a source for product usage. Support (for example Zendesk) and the accounting system add a lot for churn signals and payment status. Start with those three and only expand once the matching at customer level works.

How much revenue does a SaaS company leak on average?

There is no reliable industry average for that. The commonly cited estimate for revenue leakage in general is 1 to 5 percent of revenue. In SaaS it depends mainly on how much custom work you sell, whether you bill on usage and how often contracts change mid-term.

Is Revenue Intelligence only for large SaaS companies?

No. It is precisely at a company with a few hundred customers and a small finance team that discrepancies are likely to be left unaddressed, because nobody has time to reconcile every month. Size mainly determines whether a spreadsheet will do or whether you need software.

Where do you start if CRM and billing cannot be connected?

With a shared customer ID. As long as you cannot say per customer which CRM record belongs to which billing subscription, you cannot compare anything. Setting that up is dull work, but it is the foundation for everything above.