Operator analysis · Independent report · 2026

State of B2B Monetization 2026 — The Six Takeaways That Matter

Kyle Poyar dropped the 2026 State of B2B SaaS and AI Monetization Report on May 13 2026 — a 37-page benchmark surveying 230 B2B software and AI companies across April – May 2026. This page is the operator read on the six findings that actually change buying decisions, with attribution back to the source on every data point and StackSwap commentary on what each one means for founders, RevOps leaders, and procurement teams.

The six takeaways

Each takeaway below pairs the survey stat with operator commentary on what it actually means for the decisions on your desk this quarter.

#1. Hybrid pricing has crossed the chasm

37% of B2B companies now run hybrid pricing — up from 25% twelve months ago.

Hybrid pricing (typically per-seat + AI consumption, or flat-fee + usage overages) is now the most common pricing shape in B2B software. Early-stage <$5M ARR companies still skew to flat-fee subscriptions at 37%; large >$150M ARR companies cling to per-seat at 29%. Investor preference has flipped — 35% favor hybrid, 26% favor outcome-based, only 5% still want pure seat-based.

Operator read: The honest read: if you launched in the last 18 months and you are still selling pure per-seat, you are pricing yourself into the wrong investor pitch deck. Hybrid is not a trend — it is the new default, and the "complicated to explain" objection is real but losable. The hard part is not the model design, it is the sales-team retraining and the comp-plan rewrite that has to follow it.

#2. Everyone is panicking about expansion revenue

"Not enough expansion revenue" is the #1 complaint across every pricing model.

Seat-based companies fear their model is structurally not future-proof. Usage and outcome-based companies struggle to forecast revenue. Hybrid users report customers cannot explain their bill back to procurement. The proven expansion plays from the report: premium editions at 50-100% price premium, segment-targeted add-ons, and consumption limits on the top 10% of power users (who routinely drive 70%+ of token consumption).

Operator read: The "no expansion revenue" complaint is usually a packaging problem disguised as a pricing problem. You do not have an expansion gap — you have a value-metric gap. If your account managers cannot point to a measurable customer outcome that scales with what you charge for, expansion is going to feel like extraction every time. Fix the value metric, then the expansion conversation gets easy.

#3. Pricing reshuffles are accelerating — and the biggest companies are leading

75% of software companies changed pricing or packaging in the last year.

In 2026 alone: Clay introduced dual-track monetization (March), HubSpot announced outcome-based pricing for Breeze AI agents and cut Fin pricing 50% (April), Anthropic lowered enterprise seat prices and shifted aggressively to usage-based, SAP committed to AI consumption pricing, and Salesforce, OpenAI, GitHub, Figma, and Canva all made major pricing changes. AI credits adoption sits at 29% (up 126% YoY) with 33% planning to introduce them in the next 6-12 months — including roughly half of all $50M+ ARR companies.

Operator read: Pricing changes used to be a once-every-3-years event. They are now an annual ritual at the enterprise level, and quarterly at AI-native companies. The implication for buyers is brutal: any vendor contract you signed in 2024 is on a pricing model that no longer matches the vendor's 2026 economics, which means renewal negotiations are going to look nothing like the original deal. If you are running procurement, build the assumption that "renewal = re-price" into your forecasting now.

#4. AI margins are 50%, not 80% — and pricing teams know it

Median AI gross margin target: 50%. SaaS legacy benchmark: 70-80%+.

Only 12% of AI companies are targeting SaaS-like 80%+ gross margins. A similar share is consciously targeting 20% or lower (PostHog being the cited example). Internal costs and margins were the #1 pricing factor cited by 54% of respondents — beating market/competitive pricing (36%) and human productivity gains (30%). The takeaway: AI pricing is being engineered around cost-of-goods-sold first, value second.

Operator read: This is the most under-discussed number in the report. The entire B2B SaaS pricing playbook assumed software margins were 80%+ and you priced on value, not COGS. AI inverts that: when 50% of every dollar goes to inference costs, you cannot price purely on value or you will go broke at scale. The companies winning here are the ones treating margin engineering as a product surface — caching, model routing, smaller-model fallbacks — not just a finance team problem.

#5. AI is eating the software budget — not new budget

70% of respondents said AI spend comes out of customers' tech/software budgets, not new budget.

SaaS companies introducing AI features cannibalize tech budgets in 75% of cases — they are taking from their own future renewal pool. AI-native services companies show a different pattern: 35% tap services budgets and 15% occasionally access headcount budgets, positioning themselves as Accenture alternatives with outcome-based pricing rather than as SaaS expansions.

Operator read: If you are a SaaS vendor selling AI add-ons to existing customers, you are mostly trading dollars from one of your own SKUs to another, not growing the pie. The companies actually expanding the addressable budget are the AI-native services players going after the consulting line item, not the software line item. For B2B SaaS founders, the strategic question is whether your AI feature should sell into your existing customer's software budget (margin-dilutive trade) or whether you can credibly reposition it as a services replacement (margin-accretive expansion).

#6. Multi-model pricing is becoming a sales weapon

29% of B2B companies now offer customers a choice between multiple pricing models — up from 21% last year.

Decagon offers per-conversation or per-resolution. Salesforce Agentforce runs four concurrent models: per-conversation, flexible credit, per-user add-ons, and a flat-rate Agentic Enterprise License Agreement. Multi-model pricing is most common among usage-based and outcome-based companies — they use it as a sales-cycle weapon, letting buyers pick the model that wins the internal procurement debate.

Operator read: Letting customers pick their own pricing model used to be a forecasting nightmare. With AI, it is becoming a deal-closer. The Salesforce Agentforce four-model setup is the canonical example: by the time procurement has finished arguing about which model is "right," the AE has already won the decision because every option ends in "yes, we buy." For founders, the trade is forecasting volatility in exchange for win-rate lift. At sub-$20M ARR, the volatility usually wins. At $50M+, the win-rate lift starts to dominate.

The 2026 vendor pricing reshuffles

Every major vendor on this list changed pricing in 2026. The pattern is no longer “once-every-three-years repricing” — it is now an annual ritual at enterprise scale and quarterly at AI-native companies. If you are running procurement, this table is your renewal-risk register.

VendorWhen2026 pricing move
ClayMarch 2026Introduced dual-track monetization — separated platform value from token costs so customers pay for the workflow surface and the underlying token spend independently.
HubSpotApril 2026Announced outcome-based pricing for Breeze AI agents and cut Fin pricing 50% — the largest single-month repricing of an AI agent product in the B2B SaaS category to date.
AnthropicEarly 2026Lowered enterprise seat prices and shifted aggressively to usage-based pricing — a pure model swap from per-seat to token consumption at the enterprise tier.
SAPEarly 2026Announced a strategic shift toward AI consumption pricing for the agent portfolio — the largest legacy enterprise software player publicly committing to consumption.
Microsoft (GitHub Copilot)June 2026 launchLaunching AI credits for GitHub Copilot. CEO Satya Nadella reframed seats as "just entitlement to some consumption" — explicit signal that the seat metric is now a budget envelope, not a value metric.
DecagonOngoingOffers per-conversation OR per-resolution pricing — the canonical "let the buyer pick" multi-model example for AI-native support agents.
Salesforce (Agentforce)OngoingRuns four concurrent pricing models: per-conversation, flexible credit, per-user add-ons, and a flat-rate Agentic Enterprise License Agreement. Most aggressive multi-model deployment in the category.
OpenAI / GitHub / Figma / CanvaEarly 2026All four made major pricing changes in early 2026 — collectively signaling that pricing reshuffles are now standard practice at scale, not a once-in-a-cycle event.

Investor preference by pricing model

The investor preference data is the most underappreciated finding in the report. The historical “investors hate variable pricing” mantra has fully inverted — 35% now favor hybrid, 26% favor outcome-based, and only 5% still favor pure seat-based. If you are pitching investors on pure per-seat in 2026, you are pitching into a preference distribution that no longer exists.

Pricing modelInvestor preference
Hybrid35%
Outcome-based26%
Usage-based24%
Flat-fee10%
Seat-based5%

The CAMP framework for outcome-based pricing

Outcome-based pricing is the model investors most want (26% preference) and the one most companies cannot actually execute. The CAMP framework — Consistency, Attribution, Measurability, Predictability — is the four-test rubric for whether your outcome candidate is monetizable. If any one of the four fails, the model collapses at the invoice-dispute step.

C
Consistency
The outcome you charge for must produce a repeatable result across customers — not "sometimes works." If outcome consistency varies wildly by customer, you cannot anchor pricing on it.
A
Attribution
You must be able to prove your product caused the outcome, not just correlated with it. Without clean attribution, customers will dispute every invoice.
M
Measurability
The outcome must be quantifiable in a way both parties accept. Vague outcomes ("better customer satisfaction") fail this test; specific outcomes ("resolved tickets without human escalation") pass.
P
Predictability
The customer must be able to forecast their bill within a reasonable range. Pure pay-per-outcome with high variance creates procurement panic — predictability is what makes outcome-based pricing palatable to buyers.

Survey respondent demographics

230 respondents, well-distributed across company stage, ACV band, and product mix — which means the takeaways above generalize across early-stage startups, mid-market, and enterprise. The 43% “SaaS + AI hybrid” product mix is especially telling: most respondents are not pure-play AI companies, they are SaaS companies navigating the AI margin problem in real time.

Respondent ARR

<$1M22%
$1-20M28%
$20-150M24%
>$150M25%

Average customer ACV

<$5K28%
$5-25K27%
$25-100K30%
>$100K16%

Product mix

Mostly SaaS32%
AI-native15%
SaaS + AI hybrid43%
Other10%

The four quotes worth keeping

Tech investors used to hate variable pricing. They're pushing for it in the age of AI.
Kyle Poyar, 2026 State of B2B Monetization Report
Remember there's no such thing as a perfect pricing model.
Kyle Poyar, 2026 State of B2B Monetization Report
AI credits are great for vendors. They can become a nightmare for customers, especially once teams have to manage different credit models across dozens of their vendors.
Kyle Poyar, 2026 State of B2B Monetization Report
When products look like a commodity, companies don't have much pricing power. Credits or tokens are the ultimate commodity pricing metric for AI.
Kyle Poyar, 2026 State of B2B Monetization Report

What this means for buyers right now

Three operator moves the data makes obvious for anyone responsible for software spend in the next two quarters.

  1. Build “renewal = re-price” into your forecasting. 75% of vendors changed pricing in the last year. Your 2024 contracts are no longer the deal the vendor would sign with you today. Re-run TCO against alternatives 90 days before every renewal, not 30 days after the renewal notice.
  2. Audit AI add-ons separately from base SKUs. 70% of AI spend cannibalizes existing tech budgets — which means most AI features you bought from existing vendors are funded by reducing usage of the base product. If you do not see that trade explicitly on the renewal worksheet, the vendor is taking a margin-positive trade and you are taking a margin-negative one.
  3. Pressure-test outcome-based offers against CAMP. The wave of outcome-based AI pricing (Decagon, HubSpot Breeze, Agentforce) is real. Before signing one, run the offer through Consistency / Attribution / Measurability / Predictability. If any of the four fails, you will spend the contract life disputing invoices.

What this means for founders right now

Three operator moves the data makes obvious for B2B SaaS and AI founders setting pricing for 2026.

  1. If you are still on pure per-seat, plan the migration. Only 5% of investors prefer pure seat-based. 35% prefer hybrid. The transition cost is real (sales retraining, comp plan rewrite, customer comms) but the status-quo cost is becoming structural — you will lose enterprise deals to hybrid-priced competitors.
  2. Treat margin engineering as a product surface. 50% AI gross margin is the new median, not the floor. The teams winning here ship caching, model routing, smaller-model fallbacks, and cost telemetry as part of the product roadmap — not as a finance team problem. If your AI feature has no margin instrumentation in production, you are flying blind on the most important economics in the business.
  3. Consider multi-model pricing as a sales weapon. 29% of B2B companies now offer customers a choice. Decagon and Salesforce Agentforce are the canonical examples. The trade is forecasting volatility for win-rate lift — at sub-$20M ARR the volatility usually wins; at $50M+ the win-rate lift starts to dominate. Worth modeling for any AI-native motion above $5M ARR.

FAQ

The 2026 State of B2B SaaS and AI Monetization Report is a 37-page benchmark published by Kyle Poyar at Growth Unhinged on May 13 2026. It surveys 230 B2B software and AI companies across April-May 2026 on pricing models, packaging, AI margins, expansion revenue, and the major pricing reshuffles happening across the category. Full report: https://www.growthunhinged.com/p/the-state-of-b2b-monetization-in-2026.

37% of B2B software companies now run hybrid pricing (typically per-seat + AI consumption, or flat-fee + usage overages) — up from 25% twelve months prior. Hybrid is now the single most common pricing shape in the category. Early-stage <$5M ARR companies still skew to flat-fee subscriptions (37%); large >$150M ARR companies cling to per-seat (29%). Investor preference for hybrid sits at 35%, with outcome-based at 26% and pure seat-based at just 5%.

75% — three in four B2B software companies changed pricing or packaging in the trailing twelve months. Notably, the largest companies are now making the most pricing changes, reversing the historical pattern. 2026 saw major repricing announcements from Clay (dual-track monetization, March), HubSpot (outcome-based Breeze AI + 50% Fin cut, April), Anthropic, SAP, Salesforce, OpenAI, GitHub, Figma, and Canva.

The median AI gross margin target is 50%, versus the 70-80%+ benchmark for legacy SaaS. Only 12% of AI companies target SaaS-like 80%+ margins; a similar percentage consciously target 20% or lower (PostHog is cited as the notable example). Internal costs and margins were the #1 pricing factor cited by 54% of respondents — beating market/competitive pricing (36%) and human productivity gains (30%).

70% of respondents said AI spend comes out of customers' tech/software budgets — not new budget. SaaS companies introducing AI features cannibalize tech budgets in 75% of cases. AI-native services companies show a different pattern: 35% tap services budgets and 15% occasionally access headcount budgets, positioning themselves as Accenture alternatives with outcome-based pricing rather than as SaaS expansions.

CAMP stands for Consistency, Attribution, Measurability, Predictability — the four preconditions for outcome-based pricing to work. Consistency: the outcome reliably reproduces across customers. Attribution: you can prove your product caused the outcome. Measurability: the outcome is quantifiable in a way both parties accept. Predictability: customers can forecast their bill within a reasonable range. Pricing teams should test outcome candidates against all four before committing to a model.

29% of B2B companies now offer customers a choice between multiple pricing models — up from 21% last year. Decagon offers per-conversation or per-resolution. Salesforce Agentforce runs four concurrent models: per-conversation, flexible credit, per-user add-ons, and a flat-rate Agentic Enterprise License Agreement. Multi-model pricing is being used as a sales-cycle weapon — letting buyers pick the model that wins their internal procurement debate.

29% of B2B companies currently use AI credits, up 126% year-over-year. 33% plan to introduce AI credits in the next 6-12 months — including roughly half of all $50M+ ARR companies. Microsoft is launching AI credits for GitHub Copilot in June 2026, with Satya Nadella reframing seats as "just entitlement to some consumption" — explicit signal that seats are becoming budget envelopes, not value metrics.

Build "renewal = re-price" into your procurement forecasting. Any contract signed in 2024 was priced against the vendor's 2024 economics — those economics no longer match 2026, which means renewal negotiations are going to look nothing like the original deal. Audit existing vendor spend against current pricing pages before renewal, run a TCO comparison against alternatives, and budget for either a higher renewal or a vendor swap. StackScan applies this rubric to your full stack in 60 seconds.

The full report is at https://www.growthunhinged.com/p/the-state-of-b2b-monetization-in-2026. Premium subscribers get the 37-page report plus a subscriber-only AMA. There is also a related Metronome webinar (Scott Woody, "How Agents Are Changing Monetization," May 14 2026) referenced in the same release.

Related reading

Source: 2026 State of B2B SaaS and AI Monetization Report by Kyle Poyar, published at Growth Unhinged on 2026-05-13. Original report: https://www.growthunhinged.com/p/the-state-of-b2b-monetization-in-2026. Stats quoted with attribution; commentary and operator analysis are StackSwap's own framing. Kyle Poyar, "2026 State of B2B SaaS and AI Monetization Report," Growth Unhinged, May 13 2026. https://www.growthunhinged.com/p/the-state-of-b2b-monetization-in-2026

Canonical URL: https://stackswap.ai/state-of-b2b-monetization-2026