← All articles

Strategy

Your usage pricing won't fail with customers. It'll fail in the comp plan.

With seats, the deal ends at signature. With usage pricing, it starts there — and most sales teams are still paid for the old ending.

On March 7, 2024, MongoDB's CFO told investors something you rarely hear on an earnings call: "in fiscal 2024, we changed our sales incentive structure to reduce the importance of upfront commitments."

The numbers around that sentence explain why it mattered. MongoDB had recognized over $40 million in fiscal 2024 from unused Atlas commitments — customers paying for capacity they'd committed to and never consumed. For fiscal 2025, it expected "close to zero revenue from unused Atlas commitments." In its results release, the company flagged "over $80 million of FY24 revenue, related to multi-year term licenses and unused Atlas commitments" as a headwind for the year ahead.

Read those numbers side by side and the lesson is hard to miss. When reps are rewarded for the size of the commitment, some commitments outrun what customers actually use. The revenue looks real — once. Then the customer doesn't grow into it, and it doesn't come back.

The reversal nobody trains for

Moving from seats to usage changes the sales job more than it changes the pricing page.

Selling seatsSelling usage
Quote a flat numberSell a starting commitment plus growth
Discount seats to closeExplain credits, overages, and resets
Success is contract value at signatureSuccess is consumption after signature
The deal ends when it's signedThe deal starts when it's signed

A rep who spent ten years learning to quote, discount, and close now has to forecast a customer's usage curve and stay involved after the ink dries. That's not resistance. It's a different job.

Even Salesforce is still looking for the sellable unit

Look at how Agentforce pricing evolved.

  • September 2024 — $2 per conversation.
  • May 15, 2025 — Flex Credits: one action consumes 20 credits, or $0.10, sold at $500 per 100,000 credits. Salesforce's stated reason: "Organizations are increasingly seeking a pricing model aligned to how AI agents deliver business outcomes and create value."
  • June 17, 2025 — add-ons "starting at $125 per user per month," which Salesforce's pricing page describes as "Unmetered Agentforce usage for employees."
  • October 2025 — the Agentic Enterprise License Agreement, announced at Dreamforce: a broad license for large deployments. Salesforce's CRO, Miguel Milano, on why: "Predictability of cost was very important for CEOs." Marc Benioff has since described it as the preferred licensing path for large deployments, according to Techstrong.ai.

All of these models are still on sale. SaaStr's Jason Lemkin read it this way: "It's not confusion — it's a hedge."

Notice the direction. The per-action credit is the unit that maps most precisely to what an agent does. The license Salesforce steers its biggest deals toward is the one a CEO can budget — and a rep can close. The pricing unit and the sales motion turned out to be the same decision.

Where the comp plan breaks

Pay reps only on the commitment, and you invite commitments bigger than real usage — the pattern behind MongoDB's numbers. Pay them only on consumption, and you pay them for something they don't control: the customer's adoption curve. Your best hunters stop hunting.

Snowflake landed in between. Its CFO, Mike Scarpelli: "we had to change our compensation model so it wasn't just paying on the booking." Snowflake's revenue operations lead described a blend that shifts as a territory matures — roughly 70% bookings and 30% consumption in new territories, moving to about 30/70 in mature ones. Their principle: "Pay should be based on the value that a sales rep helps a customer realize by using your solution."

The structure I recommend in the playbook follows the same logic:

  • Pay on committed value at signature. The rep controlled the commitment.
  • Add a consumption bonus when the account exceeds its commitment within 12 months. Now the rep wants commitments the customer will grow into — not ones that age into unused balance.
  • Hand steady-state growth to customer success once usage stabilizes. Keep step-change expansion — a new department, a new budget owner — with the account executive.

With seats, the risky moment is the close. With usage pricing, it's the months after it.

Why this lands on the PM's desk

If you're the PM moving a product from seats to usage, the comp plan isn't someone else's problem. Your pricing unit decides what a rep can forecast, what a buyer can budget, and which contract shapes make sense: a prepaid credit pool, a commitment plus overage, pay-as-you-go, or pay per outcome. Each one gets sold, forecast, and paid differently.

The upside is real when it works. OpenView's 2021 survey of 600 SaaS companies found that, among top-quartile performers, usage-based companies reached 122% net dollar retention, against 109% for subscription-only companies. Expansion is the whole point of usage pricing. But expansion after signature only happens if someone is paid to make it happen.

Before you change the pricing page, change the comp plan. Otherwise the pricing page won't matter.

If your best rep sold a usage deal tomorrow, would they earn more for a commitment the customer grows into — or for one the customer never uses?

Free companion

The CRO's Guide to Selling Consumption

How to retrain a seat-based sales team for usage pricing, in a single PDF — the human-cost pitch, the four contract structures every rep must know, a comp model without clawback traps, the three expansion plays, and a 90-day retraining plan.

Get the guide →

This is one piece of a longer framework I teach in Chapter 5 of Product Strategy in the AI Era — including the Agentic ELA: trading unused seats for AI credits, so you protect contract value when the seat stops being the unit.

Sources

  • MongoDB Q4 FY2024 earnings call transcript (March 7, 2024): marketbeat.com
  • MongoDB Q4 FY2024 results release (SEC filing, March 7, 2024): sec.gov
  • Bloomberg Law, "Salesforce to charge $2 per conversation" (September 17, 2024): bloomberglaw.com
  • Salesforce, Agentforce flexible pricing (May 15, 2025): salesforce.com
  • Salesforce, Agentforce add-ons (June 17, 2025): salesforce.com
  • Salesforce, Agentforce pricing: salesforce.com
  • Diginomica, Dreamforce 25 and the Agentforce bill (October 16, 2025): diginomica.com
  • Techstrong.ai, Salesforce's agent pricing moves toward seats (December 15, 2025): techstrong.ai
  • SaaStr, Salesforce's three Agentforce models (February 17, 2026): saastr.com
  • Stripe, customer story with Snowflake CFO Mike Scarpelli: stripe.com
  • Snowflake, "Sales compensation in a consumption pricing world" (March 24, 2022): snowflake.com
  • CFO Dive, OpenView usage-based pricing data (November 4, 2021): cfodive.com

Newsletter

The Strategic AI Corner

One idea on product strategy in the AI era — weekly. No spam, unsubscribe anytime.

Free. Unsubscribe anytime.

More articles

📊
Strategy

Figma gave customers three months to learn its AI credits. One still ended up at 209% of the limit.

Measuring usage before you charge for it is the right move. Figma did it by the book — and its forum still shows exactly where warm-up periods break.

Read →
📉
Strategy

Your engineers just made your biggest pricing decision

Two AI products can charge the same price and keep very different shares of it. The difference gets decided in the architecture — and nobody calls it a pricing decision.

Read →
📋
Strategy

Your customer's procurement team now has a checklist for your AI pricing

Buyers have spent two years absorbing AI price increases. Now they're organizing. Here's what they'll ask — and how to design pricing that passes before they do.

Read →
🧾
Craft

Stripe didn't buy a billing company. It bought the line between product and business.

Credits look like a finance detail. Each one carries a dozen product decisions — and in 2025, the companies that got them wrong found out in public.

Read →
🧭
Strategy

Anthropic pricing and model — a lesson for all respected SaaS

Anthropic's top plan unlocks four things, and not one of them is a feature. A masterclass in AI-era segmentation — and what it means for your own pricing.

Read →
📝
Craft

Your PRD now has three readers — and one of them is an AI

PRDs didn't die. They got a new audience — and old habits don't survive it.

Read →
💰
Strategy

Pricing is the feature you forgot to ship

Why pricing decisions belong in the product roadmap, not in finance.

Read →
🎯
Positioning

Every AI feature you add is a positioning decision you didn't make

The hidden cost of adding AI features without a positioning frame.

Read →