Stripe Puts Hybrid Pricing at One in Six Scaled Users
Metronome's CEO says roughly one in six Stripe users past key revenue milestones were using or rolling out seat-plus-usage pricing by August. Metronome's hybrid support had sat largely unused for about 18 months.
October 5, 2026
Scott Woody spent most of his October 1 post on Stripe's blog arguing against a pricing model his own product makes easy to run. Woody is the CEO of Metronome, the usage-billing company Stripe finished acquiring in January. His target is token billing, which charges customers for the model tokens they burn plus a markup. Woody says it turns a product into "a commodity markup on top of an actual commodity."
The number most people will quote sits near the end. Metronome built support for hybrid pricing, a mix of seats and usage, about three years ago, and its customers mostly ignored it for around 18 months. According to Woody, as of August 2026, roughly one in six Stripe users that had crossed key revenue milestones were using or rolling out hybrid pricing. Stripe doesn't say how many users that is or define the milestones. It also sells the billing software, so this is a vendor's adoption signal rather than a market benchmark.
A second count points the same way from a different base. Kyle Poyar's 2026 survey of 230 software companies found 37% running hybrid pricing, up from the 25% who said they were hybrid a year earlier. Poyar's respondents opt in, which may explain why his share is about double Stripe's. His newsletter is also sponsored by Metronome, so neither number is independent of the billing vendor.
Revenera's 2027 Monetization Monitor, released October 14, is a third count, and it doesn't come from Metronome or a newsletter Metronome sponsors. Its survey of 415 product and business leaders at software companies, fielded in June and July, found 37% monetizing AI with a hybrid of subscription plus usage or outcome pricing, the same share Poyar found for hybrid pricing overall. Revenera sells software monetization technology, so this is vendor research too, though it says only 8% of respondents are its customers. The same survey found 41% still rely on manual processes or engineering effort to collect usage data, the data that both a consumption-weighted comp plan and the capped usage meter vendors are attaching to seats run on.
Stripe added a fourth count on October 19, this time of plans rather than adoption. Its survey of 200 C-suite and director-level leaders at SaaS platforms in North America and Europe, fielded in March with half the sample on Stripe, found one in four plan to keep per-seat fees and charge separately for agent work on credits, usage or outcomes, and another 7% plan to price agents like seats. Respondents expect seats to fall from 16% to 11% of platforms' primary pricing model by 2027, and 44% plan several pricing changes in the next 12 months. These are expectations from a sample the billing vendor picked, but a quarter of those roadmaps now include the seat-plus-variable deal this piece's comp advice is written for.
Hybrid pricing moves the hard problem from the invoice to the commission plan
Billing is the part of hybrid pricing that's already solved, and Stripe now owns that fix. A hybrid deal produces two revenue lines with different shapes. One is a fee the customer signs for on day one. The other is a usage line that arrives later, if the customer adopts. The buyer can't predict that second line any better than the vendor can, so procurement may ask for a cap or a prepaid block before signing. A prepaid block turns part of the usage line back into a commitment. A bookings-based commission plan pays on commitments, because a signed contract is what the CRM can see on close day. On a hybrid deal, most of the rep's variable pay should follow consumed revenue instead, even though that makes both the paycheck and the CFO's forecast lumpier.
Woody's preferred fix makes the question sharper. He wants companies to show customers unified credits, a shared balance that gets drawn down as the product does work. The invoice then lists outputs like enriched records or generated images instead of models and markups. That's a better invoice. But a prepaid credit block is a commitment under another name. A plan that pays the rep when the block is sold pays for revenue that may never be consumed. Linear, for one, sells AI credits as a prepaid balance on top of its seat plans, and the funds expire 12 months after purchase. It's the same company we noted bills autonomous Coding Sessions on a separate line while keeping per-seat pricing for people. Expiring credits are where unused-commitment revenue comes from.
MongoDB already showed what paying on commitments buys
MongoDB is the public case. In fiscal 2024 the company changed its sales incentive structure to reduce the importance of upfront commitments, then-CFO Michael Gordon said on its March 2024 earnings call. MongoDB then saw far fewer upfront commitments, and the revenue from commitments customers never used went with them. Gordon said MongoDB expected close to zero revenue from unused Atlas commitments in fiscal 2025, against more than $40 million in fiscal 2024.
That $40 million is what paying on commitments looks like once it's itemized. Under the old plan, the rep was paid and the company booked high-margin revenue for a while. Customers paid for capacity they didn't use, which is the kind of line a buyer remembers at renewal.
Snowflake discloses the split openly. Its fiscal 2026 annual report says part of the sales force's commission is earned on customers' consumption and part on the origination, expansion or renewal of contracts. A hybrid revenue model with a hybrid comp plan already sits in a public filing, so nobody has to invent one.
Reps can't control usage, and finance loses its easiest number
A rep can't control consumption. Usage can fall because the customer's own market had a bad quarter, and paying on burn means the rep absorbs a drop nobody in the deal caused. It also pulls reps toward accounts where usage already flows and away from new logos.
Finance loses something real too. A signed commitment is the easiest forecast a CFO gets. Gordon added the unused commitments to about $40 million of extra multiyear license revenue. Together that was roughly $80 million of very high-margin revenue that wouldn't repeat, and MongoDB's fiscal 2025 guidance had to absorb it. In Poyar's survey, the companies on usage and outcome pricing say revenue is hard to forecast.
That objection gets the forecast right. Still, part of the predictability a commitment-paid plan hands the CFO is revenue the customer won't use. A buyer watching credits expire has every reason to negotiate it away. The better plan pays enough at signature that closing still matters, then puts the larger share on consumption, trued up over the contract year.
What this means for CROs and RevOps leaders
Split every hybrid quota into a commit line and a consumption line before the plan is published. Decide the weighting on purpose rather than defaulting to bookings because the CRM already tracks them. Snowflake's filing is a reasonable reference to bring into the room.
Have the deal desk size credit blocks to the customer's expected usage. A block sized to the largest number procurement will approve flatters this year's bookings and sets up the renewal fight.
Give the CFO a forecast built on burndown before the first procurement call. It should show how fast each cohort of accounts draws down its credits and how much of the commit is left at month nine. That's the forecasting answer hybrid pricing needs, and it comes from billing data the company already has.
Keep the invoice on outputs, as Woody argues. A rep can sell a resolved ticket or an enriched record but can't sell a markup on someone else's model.
My bet is that by the end of 2027 the standard hybrid comp plan pays the majority of variable compensation on consumption, with signature pay shrunk to something closer to a closing bonus. The commission language in the next round of SaaS annual reports will show whether Snowflake's split became the norm.
But a prepaid credit block is a commitment under another name. A plan that pays the rep when the block is sold pays for revenue that may never be consumed.
Split every hybrid quota into a commit line and a consumption line before the plan is published, and size credit blocks to the customer's expected usage rather than the largest number procurement will approve.
