The billing layer problem: Why consumption pricing breaks most companies before they know it

Article    July 15, 2026
The billing layer problem: Why consumption pricing breaks most companies before they know it
SHARE
SHARE
BOTTOM LINE UPFRONT

The shift to consumption pricing is accelerating, and Salesforce’s acquisitions of m3ter and Fin show that billing infrastructure is becoming a competitive advantage. PE-backed CFOs that modernize the billing layer can capture more revenue, improve financial accuracy, and tell a stronger value creation story at exit.

Salesforce bought the meter before it bought the thing worth metering. In the span of two weeks, it acquired m3ter, a metering and rating platform for consumption-based monetization, then Fin, the AI customer service company (formerly Intercom) whose AI agent autonomously resolves 76% of support volume end-to-end. The sequencing was deliberate: you can only charge for an outcome you can measure.

Salesforce framed both acquisitions around giving customers more flexibility to monetize as AI pushes pricing toward consumption, and analysts covering the deals were blunt about why: most enterprises are still running billing infrastructure built for seats, even as their products grow more usage-heavy by the quarter.

There’s also a more self-interested read tucked inside the timing. Salesforce already needs to meter Agentforce usage, and the Fin acquisition is about to increase that metering need significantly. Since Salesforce had already partnered with m3ter, buying them outright to meet that need was the logical next step. One is the plumbing to bill for agentic work; the other is the agent that does the work.

Either way, the message for PE-backed CFOs is this: the dominant CRM vendor is retooling its entire revenue platform around consumption pricing. Most companies aren’t ready for it, and the gap is already costing money.

The shift from flat subscription pricing to consumption-based models is already here

Pricing has moved toward outcomes: AI products billed per interaction, infrastructure billed per compute hour, professional services billed per milestone. But most companies’ billing infrastructure is still built for the pricing model they’ve already left behind.

Legacy CPQ and billing systems were designed around the signed contract; they handle fixed-amount invoicing reliably, but that’s roughly where their usefulness ends. When consumption enters the picture, the cracks appear fast: usage tracked in spreadsheets outside the core system, a monthly close that drifts while finance reconciles by hand, charges dropped because the rating logic can’t keep pace. The result is revenue leakage: companies lose between 1–5% of realized EBITDA annually, and over 40% experience some form of it. For a portfolio company underwriting to an aggressive growth plan, that’s significant margin walking out the door.

What goes wrong when you bolt it onto legacy CPQ

The instinct is to extend what already exists. It rarely works cleanly. Here’s where it breaks:

  1. Metering and rating aren’t native: Legacy CPQ was built to price a configuration at a point in time, not to capture and rate a continuous stream of usage events. Bolting on a metering layer means reconciling two systems that weren’t designed to talk to each other, and every gap between them is a potential dropped charge.
  2. Mid-cycle changes compound the problem: Consumption models change in-flight: upsells, overages, downgrades, proration. A system built for fixed monthly charges either fails to handle these cleanly or forces a manual credit-and-rebill cycle that erodes confidence in the numbers, and creates diligence risk before a buyer ever arrives.
  3. Revenue recognition gets harder: Usage revenue introduces variable consideration and timing rules under ASC 606 that a fixed subscription schedule never had to handle. Purpose-built platforms like RightRev exist precisely for this, automating allocation and timing natively rather than as a downstream reconciliation exercise. When that work happens outside the system of record, the close slows, the audit trail weakens, and the risk of restatement climbs.
  4. AI economics expose the gap fastest: AI products carry a cost that scales with every interaction. When metering can’t keep pace with usage volume, margin erosion accelerates exactly as the product gains traction, the worst possible moment to lose billing precision.

Why RCB changes the equation

Revenue Cloud Billing was built for this rather than retrofitted for it. Native metering, rated usage, and flexible pricing models aren’t bolted-on modules, they’re core to how the platform was architected from day one. Usage events flow directly into rating and invoicing without a side system translating between them. Mid-cycle amendments, like upsells, overages, and downgrades, process in real time instead of triggering a manual credit-and-rebill cycle. Revenue recognition runs inside the system of record, so ASC 606’s variable consideration and timing rules are handled natively rather than reconciled after the fact. Usage events trace cleanly to the invoice, giving finance an audit trail instead of a spreadsheet.

And RCB already has advanced native usage-metering capabilities today. The m3ter acquisition doesn’t introduce metering to the platform; it sharpens what’s already there, adding more sophisticated rating logic for complex, high-volume usage patterns as consumption models get more intricate.

For companies running consumption pricing, or planning to, the difference between a billing layer that was built for it and one that was retrofitted for it shows up in three places: margin capture during the hold, close quality at audit, and the revenue story a buyer can actually follow in the data room.

The question worth asking now

Salesforce just spent two acquisitions and several billion dollars signaling where enterprise monetization is heading: agents doing the work, and infrastructure metering what it’s worth. That’s not a hedge, it’s a bet on the direction the entire market is moving.

The business has already committed to the pricing model. The only open question left for a PE-backed CFO is whether the billing infrastructure gets fixed on their timeline, during the hold, when there’s still time to capture the margin, or on a buyer’s timeline, in diligence, when the gap becomes their problem to price into the deal.

 

FAQ

Why did Salesforce acquire m3ter and Fin within two weeks of each other?

The sequencing reflects a deliberate bet on where enterprise monetization is heading. m3ter brings native metering and rating for consumption-based pricing, while Fin brings an AI agent that autonomously resolves 76% of support volume end-to-end. Salesforce needs a way to measure and bill for that kind of agentic work at scale, and metering has to come before monetization. Together, the acquisitions signal that Salesforce is retooling its revenue platform around usage-based pricing rather than treating it as a niche add-on.

Why can't companies just bolt consumption billing onto their existing CPQ system?

Legacy CPQ and billing systems were built to price a fixed configuration at a single point in time, not to capture a continuous stream of usage events. When companies try to extend these systems for consumption pricing, the gaps show up quickly: metering logic that can’t keep pace with usage, mid-cycle changes like upsells and overages that require manual credit-and-rebill cycles, and revenue recognition under ASC 606 that has to be reconciled outside the system of record. Each of these gaps is a potential source of dropped charges, slower closes, and diligence risk.

How much revenue can leak from billing infrastructure that isn't built for consumption pricing?

Companies running consumption or usage-based pricing on infrastructure designed for flat subscriptions typically lose between 1-5% of realized EBITDA annually to revenue leakage, and more than 40% experience some form of it. For a PE-backed portfolio company underwriting an aggressive growth plan, that margin erosion compounds quietly until it becomes visible in a buyer’s diligence process.

What makes Revenue Cloud Billing different from a retrofitted billing solution?

Revenue Cloud Billing (RCB) was architected around native metering, rated usage, and flexible pricing from the outset rather than having those capabilities layered on afterward. Usage events flow directly into rating and invoicing, mid-cycle amendments process in real time, and revenue recognition happens inside the system of record instead of through downstream reconciliation. The m3ter acquisition builds on this foundation by adding more sophisticated rating logic for complex, high-volume usage patterns as consumption models continue to evolve.

Worried your billing infrastructure can't keep up? Let's talk.

Our contact form is currently blocked by your cookie preferences. Please change your preferences to continue.