Flat fee, banded by your scale
Billing-infrastructure and FinOps vendors have been replacing percentage-of-revenue take-rates with flat fees gated by a band of the customer's own scale, and four market 'not a percentage of your revenue' as the headline claim. Only 2 of 17 billing-tagged corpus companies still publish a percentage — but the trend is weakening, because the percentage often moves into the quote rather than disappearing.
What's happening — and why
What's happening: the vendors who sell billing software have been getting out of the business of taking a cut of their customers' billings. Instead of a percentage of revenue processed, they charge a flat monthly fee gated by a band of the customer's own scale — revenue, billings, tracked cloud spend, active contracts, billable end-customers, even website visitors.
It is a positioning move as much as a pricing one: four vendors market 'not a percentage of your revenue' as the headline claim, which only works as a differentiator because the alternative is common.
Why it is weakening rather than holding: the mechanic is not free of percentages, it just relocates them. Sequence's quoted tiers are still a monthly platform fee on projected billed revenue. Finout's flat bands are reported to land at roughly 1% of the cloud bill. Vantage keeps a savings-share on its Autopilot product while refusing one on its subscription. The percentage moves into the quote or onto the outcome rather than disappearing.
And the founding evidence reversed. Hyperline's take-rate exit — the first dated one in the corpus, on 2026-07-29 — was undone six days later when it restored $299/mo plus 0.7% of revenue processed in place of a $599 flat fee.
How it works
Evidence over time
10 supporting · 8 counter — hover or tap a point for detail, click to jump to the row.
Evidence
| Company | Date | What happened |
|---|---|---|
| Hyperline | Jul 2026 | Abandoned its own take-rate. Before: Quote to Cash at $199/mo + 0.6% of billed revenue and Quote to Cash + Usage at $299/mo + 0.7%. After: a flat Launch tier at $599/month covering up to $2M in annual revenue and up to 100 active contracts, with Growth ($2M–$20M revenue, up to 1,000 contracts) and Scale (above $20M, unlimited contracts) custom-quoted. The public "10 invoices free, no credit card" trial went with it, so there is no longer a free tier. Hyperline's own framing: the prior 0.6/0.7% "alone could add hundreds to thousands of dollars a month once billed revenue climbed into six figures." |
| Kill Bill | Jun 2026 | Explicitly anti-percentage. pricing_model reads "flat-annual-fee support (Launch/Growth/ Enterprise) — explicitly never a % of revenue," and the pricing page ships a savings calculator: at $25M ARR a customer-managed package runs ~$12k Aviate software + $38k–$63k Growth Support = ~$50k–$75k/yr, versus ~$175k for a typical competitor charging 0.7% of revenue. The open-source engine itself (Apache-2.0) is free to self-host with no meter at all; the AWS Marketplace deploy is ~$40/mo plus your own AWS cost. |
| Zenskar | Jun 2026 | Three custom-quoted tiers (Starter, Standard, Enterprise) carrying the full feature set and priced by scale and support — with a dedicated pricing-page section headed "No percentage of your revenue" and an FAQ that states flatly "We don't charge a % of your revenue," aimed at Stripe Billing, Zuora and Chargebee. No dollar figures are published; third-party estimates (~$15K–$100K/yr) are contradictory and unconfirmed. |
| Vantage | Jun 2026 | Fixed-rate tiers gated by monitored cloud spend, not a cut of it: free Starter up to $2,500/mo of tracked spend (3 users, 6-month retention), Pro up to $7,500 (5 users), Business up to $20,000 (10 users, 12-month retention), Enterprise custom for unlimited spend. Frontmatter states the platform fee is "NOT a percentage-of-spend meter." The paid dollar figures (~$30 and ~$200/mo) are third-party reported, not published. Percentage economics survive only in the optional Autopilot for AWS Savings Plans, which earns on realized savings. |
| Sequence | Jun 2026 | Banded by the customer's ARR: a public Growth plan at $799/mo for startups under $1M ARR, then bespoke Core ($1–10M ARR) and Scale ($10M+ ARR). Add-ons are separate — the CPQ quote builder and revenue recognition start at $499/mo, so a full quote-to-revenue deployment on Growth raises the floor from $799 to roughly $1,300/mo, and rev-rec is gated to Core and above. Important dent: the bespoke tiers are explicitly "a monthly platform fee on projected billed revenue," so the percentage logic moved into the quote rather than off the price card. Re-verified 2026-07-21 with no separate SKU for the AI agent layer. |
| Maxio | Jul 2026 | Flat fee banded by monthly billings: Grow at $599/mo up to $100k in monthly billings (badged "Most popular"), quote-only Scale above that for advanced rev-rec, A/R, multi-entity and metering. Explicitly no per-seat charges. On 2026-07-23 the page slimmed to just those two plans — the free "Build" developer-sandbox column was removed as a listed plan (it still exists as a footer resource) — with prices unchanged. |
| Flexprice | Jul 2026 | The cleanest fully published band ladder in the set, gating one flat fee on multiple independent scale dimensions at once: Free (100k events/mo, up to $100K cumulative billing revenue), Build $500/mo (1M events, up to $250K cumulative OR $20K monthly billings), Scale $1,000/mo (5M events, up to $1.2M cumulative OR $100K monthly billings), plus a custom Mission Critical tier. The 2026-07-06 change added the $100K billings cap to the free tier — extending the double meter all the way down — and moved Scale's sign-up to sales-assisted ("Contact Us" instead of "Get Started") while leaving its $1,000/mo list price published. All list prices ($500 / $1,000 / Custom) unchanged; self-hosting stays free under AGPLv3. |
| Finout | Jun 2026 | Flat annual subscription tiered by committed cloud spend rather than metered on it, reported at roughly 1% of the cloud bill: Business ~$1,000/mo up to ~$500k of spend, Pro ~$2,000/mo up to ~$2M, Enterprise custom. Frontmatter is explicit that it is "not a per-usage meter." All figures are third-party reported — Finout publishes no price list. |
| Schematic | Jun 2026 | Bands on the customer's own billable end-customers: Free $0 (10 monetized subscriptions, 500K events/mo), Growth $200/mo (100 monetized subscriptions, 10M events/mo), quoted Enterprise on volume pricing bands with unlimited events. Seats are unlimited on every tier. The band edge is enforced as a hard product gate, not an overage line: cross the free monetized-subscription limit and your own customers can no longer check out to a paid plan. |
| Docket | Jun 2026 | The same "size of your business" logic outside fintech: an all-inclusive annual subscription banded on the customer's website traffic — Growth from $3K/mo up to 20,000 monthly visitors, Scale from $4K/mo for 20,000–100,000 visitors (the "most popular" tier), Enterprise custom above 100,000. No seat fees and no conversation caps; unlimited conversations and integrations are bundled into the traffic band. The flattest band curve in the set — a 5x visitor range for a 1.33x price step. |
Counterexamples
- Hyperline · Aug 2026 — THE KILL. Six days after the flat repricing that caused this trend to be logged, Hyperline put the take-rate back: Launch is $299/month PLUS 0.7% of revenue processed, replacing the flat $599/month tier captured on 2026-07-29. That lands close to the $199+0.6% / $299+0.7% structure the July move abandoned. The bands themselves are untouched — Launch still caps at $2M annual revenue or 100 active subscriptions (renamed from "active contracts"), Growth still covers $2M-$20M with up to 1,000 subscriptions, Scale still applies above $20M with unlimited subscriptions — so the customer-scale banding survived and the percentage came back inside it. The corpus's only dated exit from a percentage take-rate is now a six-day round trip, and this trend's own published kill condition ("if any of the ten restores a published take-rate") is met by the vendor that supplied its founding evidence.
- Chargebee · Aug 2026 — The banded flat tier deleted in favour of a pure percentage, and the rate raised while doing it. Chargebee retired both Starter (USD 0/mo, free for the first USD 250K of cumulative billing, then 0.75%) and the banded flat Performance tier (USD 7,188/yr, annual commitment billed monthly, for up to USD 100K billing/mo) for one plan called Flow, sold two ways: Pay-as-you-go at 0.80% of monthly billing value with a $0 platform fee, or Commit monthly at $99/mo plus 0.65%. Both include 100M usage events/mo; Enterprise Plus is quoted with 500M. Three things moved against this trend at once — the flat band is gone, the published percentage went UP from 0.75% to 0.80%, and the USD 250,000 free threshold that had been a defining feature of Chargebee's pricing since at least 2022 no longer appears, so the percentage now applies from a customer's first dollar.
- Metronome · Aug 2026 — The correction to this trend's own count, and it moved further the wrong way. Metronome published usage rates on its free-to-start Starter plan on 2026-07-14 — "$100,000 in billing volume included, billed at 0.8% after that" and "10M events included, billed at $0.04 / 1k events after that" — two weeks BEFORE this trend asserted that only Stripe Billing and Chargebee still published a percentage. As of 2026-08-04 those same bullets read "Billing volume billed at 0.8%" and "Events volume billed at $0.04/1k ingest events": the included-volume allotment is gone and the flat metered percentage is all that remains on the card. The metered rates themselves did not change. A Stripe-owned metering vendor publishing a percentage-of-billing rate with no included band is the shape this trend claimed the segment was leaving.
- Stripe Billing · Jul 2026 — The load-bearing holdout: still a pure percentage at 0.7% of Billing volume for recurring subscriptions and 0.4% per paid invoice on Invoicing Starter, layered on top of standard payment-processing fees. On 2026-07-22 Stripe *unbundled* advanced usage-based billing into Metronome (Startup: a $100,000 billing allotment + 10M usage events; Custom quoted) rather than flattening the fee — basic metering stays inside Billing via the Meters API up to 100M events/mo. The take-rate survives where it rides payment rails.
- Chargebee · Jun 2026 — Runs both mechanics at once, which is the honest middle case: Starter is free for the first USD 250K of *cumulative* billing and then 0.75% on billing, while Performance is a banded flat fee at USD 7,188/yr (USD 599/mo, annual commitment) covering up to USD 100K billing/mo, and Enterprise is quoted. No overage rate is published above the Performance cap. A percentage bolted onto a banded flat fee, not replaced by it.
- Orb · Jun 2026 — Moved the opposite way, and reversed its own marketing to do it. Through 2024-08 Orb billed purely on monthly event volume and its FAQ stated invoicing was included "without charging a percentage of billings"; by 2024-11 it had added billings — a cut of total invoice value issued through Orb — as a primary metric alongside events, plus a platform fee on Advanced and Enterprise. As of 2026-06-02 all three tiers are "Custom pricing" with no dollar amounts, so the take-rate is now entirely behind a quote. In the corpus's monetization signals Orb is named by 8 companies — tied with Metronome, and behind only Stripe Billing's 18.
- Usage AI · Jul 2026 — Percentage pricing is *expanding* in outcome-linked FinOps even as it retreats in billing software. Usage AI runs a pure savings-share — 20% of realized EC2 savings and 35% on RDS/ElastiCache/Redshift/OpenSearch per its AWS Marketplace listing — with no platform fee, no minimum and no contract, billed only when savings land. The 2026-07-23 redesign pulled the fee number off its own site entirely, leaving the Marketplace listing as the only published rate.
- Puzzle · Jul 2026 — Abandoned the band in the other direction. Puzzle gated five tiers on annual-expense bands in 2025-01 (Free under $25k, Basic $25/mo under $25k, Starter $50/mo under $100k, Pro $100/mo under $1M, Advanced custom) and used a $20k transaction-volume gate in mid-2026 — then on 2026-07-21 replaced the customer-scale gate entirely with a vendor-indexed "Starter is free for 2 months" promo against a $25/mo list price. It swapped a customer-indexed trigger for a clock, the reverse of this trend's mechanic.
Trivia
-
One number ties three vendors together: 0.7%. Kill Bill's own pricing page (2026-06-10) ships a calculator arguing that a $25M-ARR customer pays roughly $50k–$75k/yr flat (~$12k Aviate software + $38k–$63k Growth Support) versus ~$175k for "a 0.7%-of-revenue competitor" — a $110k+ gap that widens with growth. 0.7% is exactly the rate Stripe Billing still charges on Billing volume, and exactly the rate Hyperline abandoned on 2026-07-29. The cohort's marketing and its holdout are arguing over the same basis point.
-
Banding does not actually eliminate the percentage — it just makes it lumpy. Finout (2026-06-10) sells a flat annual subscription tiered by committed cloud spend that is reported to land at roughly 1% of the customer's cloud bill (~$1,000/mo up to ~$500k of spend, ~$2,000/mo up to ~$2M). Same outcome as a take-rate, delivered as a step function, with the vendor absorbing the variance inside each band.
-
Vantage (2026-06-10) runs both philosophies in one company: its frontmatter states the platform fee is "NOT a percentage-of-spend meter," with fixed tiers gated at $2,500 / $7,500 / $20,000 of monitored monthly spend — yet its optional Autopilot for AWS Savings Plans still earns a cut of realized savings. The percentage did not die; it moved off the subscription and onto the outcome, which is precisely where Usage AI's 20%-of-EC2-savings model is expanding.
For buyers
Ask where the percentage went rather than whether it is gone. A vendor marketing 'no percentage of your revenue' may still be pricing on a band of your revenue, quoting a platform fee derived from projected billings, or taking a share of a measured outcome — Vantage refuses a take-rate on its subscription and keeps a savings-share on Autopilot. Compute the effective rate: divide the flat band by your actual volume and compare it to the take-rate you were quoted elsewhere. Finout's flat bands are reported to land near 1% of the cloud bill, which is a take-rate with extra steps.
For vendors
Flat banding is a strong positioning story and a weak commitment if you cannot hold it. The corpus's clearest lesson here is Hyperline's: exiting a take-rate and restoring it six days later is worse for trust than never having exited, and it is highly visible because pricing pages are archived. If the economics require a percentage, consider banding the percentage rather than replacing it — a published range preserves the positioning without setting up a reversal.
Outlook — what to watch
The nine-vendor structure survives but the direction does not, which is why this reads `weakens`. The kill condition was that any of the ten restores a published take-rate, or that a new entrant wins share at a sub-0.5% rate the flat-fee cohort has to match. The first half has already happened twice — Hyperline restored 0.7% on 2026-08-04 and Orb had earlier run the reverse move, adding billings as a primary metric after advertising that it charged no percentage of billings. Watch whether the two remaining published percentages (Stripe Billing at 0.7%, Chargebee at 0.75%) become the floor the rest converge back toward.
Bottom line
Only 2 of 17 billing vendors still publish a percentage, and four market the absence of one. But the percentage mostly relocated into the quote or onto the outcome, and the trend's founding evidence reversed within a week — so treat 'flat fee' as positioning until you have computed the effective rate.
FAQ
Do billing vendors still charge a percentage of revenue?
Only 2 of the 17 billing-tagged corpus companies publish one — Stripe Billing at 0.7% and Chargebee at 0.75%. But publishing is not the same as not charging: Sequence's quoted tiers are a monthly platform fee on projected billed revenue, and Finout's flat bands are reported to land near 1% of the cloud bill.
Why is this trend marked as weakening?
Because its founding evidence reversed. It was logged on 2026-07-30 on the strength of Hyperline abandoning its take-rate the day before — the first dated exit in the corpus. On 2026-08-04 Hyperline restored it, replacing a $599 flat fee with $299/mo plus 0.7% of revenue processed. Chargebee then deleted its banded flat tier entirely a week later.
Is a flat fee better than a take-rate for a buyer?
Only if the band is wide relative to your growth. A flat fee banded by your scale converts into a step function: cheap inside a band, and a cliff at the boundary. Divide the band price by your actual volume to get the effective rate, then compare it to the take-rate you would otherwise pay — that is the only comparison that survives a growth year.
What would kill this trend outright?
A new entrant winning share at a sub-0.5% published rate that the flat-fee cohort has to match, or further restorations among the ten. Two restorations have already occurred, which is why the structure is retained but the direction is withdrawn.