38% Usage-Based, 61% Hybrid — SaaS Pricing Shift

38% of SaaS companies now price on usage in 2026, up from 27% in 2023, and pure per-seat deals have collapsed to just 8% of the market. That's the short answer. The longer answer is that AI agents broke the entire logic of charging by headcount, and the companies that adapted fastest are now growing twice as fast as the ones still selling seats.

I've watched this shift play out across a half-dozen portfolio companies over the past 18 months, and it isn't a pricing-page tweak — it's a rewrite of the SaaS revenue model that underpinned the last 20 years of the category. When a customer can run the same workload with 10 people instead of 50 because an AI agent absorbed the rest, a per-seat contract stops tracking value delivered. The vendors figuring that out first are pulling away from the ones still counting logins.

How Common Is Usage-Based Pricing in SaaS in 2026?

Usage-based pricing in SaaS has reached 38% adoption in 2026, up from 27% in 2023, while another 61% of vendors run a hybrid model that pairs a base subscription with a metered usage layer, according to Kyle Poyar's State of B2B Monetization survey of 230+ software companies. Pure per-seat pricing — where headcount is the only value metric — has shrunk to roughly 8% of the market, and IDC forecasts that 70% of software vendors will have refactored away from pure per-seat pricing by 2028.

The practical pattern in 2026 is "seat plus usage," not a wholesale abandonment of subscriptions: 37% of companies now name hybrid pricing as their primary structure, the single most common answer in Poyar's survey. The seat still anchors access and predictability; the metered layer — API calls, compute credits, AI agent resolutions, tokens — prices the part of the product that scales independently of who's logged in. That's a meaningfully different model than the flat annual contract that defined SaaS through the 2010s.

Why the "Seat Apocalypse" Is Killing Per-Seat SaaS Pricing

The forcing function behind usage-based pricing in SaaS is what the industry has started calling the "seat apocalypse": AI agents let a team shrink from 50 people to 10 while the underlying software workload — queries run, records processed, tickets resolved — goes up by an order of magnitude. A per-seat contract charges less as a customer gets more value from the product, which is backwards from every other pricing model in software history. Vendors that don't reprice around actual consumption are leaving money on the table exactly when their product is becoming more valuable.

It's also creating real budget pain on the buyer side. Zylo's 2026 SaaS pricing trends report found that 78% of IT leaders were hit with unexpected AI or consumption charges in the last twelve months, and 61% said they cut planned projects specifically because of price increases tied to that shift. That tension — vendors need to capture AI-driven value, buyers want predictable bills — is exactly why the hybrid seat-plus-usage model has become the dominant compromise rather than pure consumption pricing winning outright.

The 2026 SaaS Pricing Model Data: Snowflake, Datadog, and ServiceNow

The clearest evidence that usage-based pricing in SaaS outperforms is in the public company numbers. Snowflake's product revenue grew 34% year over year to $1.33 billion in its fiscal Q1 2027 (ended April 30, 2026), with net revenue retention at 126% — both driven by AI workload consumption on its credit-metered compute model. Datadog, which runs a hybrid per-host-plus-usage model, grew revenue 32% year over year to $1.006 billion in Q1 2026. ServiceNow now generates 50% of its net-new business from non-seat-based pricing, largely token-metered AI agent consumption. All three are growing at roughly double the ~14% median revenue growth rate for public SaaS companies in 2026.

Figures are Q1 2026 and fiscal Q1 2027 results blended from Snowflake and Datadog earnings releases, ServiceNow and HubSpot public disclosures, and public SaaS growth benchmarks as of July 2026. Median SaaS growth rate reflects the broader public software index, not a single peer set.

HubSpot's Pricing Pivot: What a 19% Stock Drop Signaled

HubSpot is the clearest cautionary tale in the usage-based SaaS pricing story. The stock dropped roughly 19% in a single trading day in 2026 as investors priced in the risk that a seat-anchored CRM couldn't capture the value AI agents were creating — the market effectively told HubSpot what per-seat pricing is worth in an AI-native world. HubSpot's response was to layer Breeze AI credits ($9 per 1,000, with 500-5,000 included monthly depending on plan) on top of its core seat model, then go further in April 2026 by moving its Breeze Customer Agent from $1.00 per conversation to $0.50 per resolved conversation, and its Prospecting Agent to $1.00 per recommended lead.

That per-resolution shift matters more than the headline price change: HubSpot is now charging for outcomes (a resolved ticket, a qualified lead) rather than raw activity (a conversation, an API call), which is the more defensible metering unit when a buyer is comparing cost against a human alternative. Expect more vendors to follow that outcome-based pattern rather than pure per-unit usage metering, because it's easier to defend against a CFO asking "what did we actually get for this."

What Usage-Based Pricing in SaaS Means for Founders Raising in 2026

For founders building SaaS today, the mistake I see most often is treating usage-based pricing as a growth-stage decision to make later. It's much easier to launch with a metered or hybrid model from day one than to migrate an installed base of seat-based customers onto consumption pricing after the fact — that migration is where churn spikes and sales cycles stretch, because you're asking existing customers to accept a model where their bill can go up if the product works well for them. I'd rather see an early-stage company anchor pricing to a usage metric that scales with the value the customer receives — records processed, workflows automated, tickets resolved — even if that means a smaller initial contract than a comparable per-seat deal.

For investors, net revenue retention above 120% and revenue growth in the 30%+ range — like Snowflake and Datadog are both posting — is increasingly a signal of pricing model, not just product quality. I track SaaS multiples and growth benchmarks on SaaS Valuations, and the pattern in 2026 is consistent: consumption and hybrid pricing companies are commanding both faster growth and, on average, richer revenue multiples than seat-heavy peers, because public markets have started pricing pure per-seat SaaS as a structurally slower-growing category.

How to Actually Implement Usage-Based Pricing in SaaS Without Breaking Renewals

The tactical failure mode I see most often isn't picking the wrong pricing model — it's picking the wrong metering unit. Twilio, one of the longest-running pure usage-based SaaS companies, priced per message and per API call from the start because that unit maps directly to cost-to-serve and to customer value at the same time. Snowflake's compute-credit model works for the same reason. The metering unit that fails is one a customer can't predict or control — bill shock, not usage pricing itself, is what actually drives churn. Zylo's finding that 61% of IT leaders cut planned projects over surprise consumption charges is a metering-design failure as much as a pricing-model failure.

The fix most 2026 vendors have converged on is a committed-use floor with usage-based overage: sell a base credit or seat allotment upfront so the customer has budget certainty, then meter anything beyond it. That's exactly the HubSpot Breeze structure (500-5,000 included credits per plan, $9 per 1,000 beyond that) and it's why 61% hybrid adoption has outpaced pure usage-based growth — buyers will accept variable pricing on the margin far more readily than they'll accept an unbounded bill. If you're a founder building this into a product from scratch, model the committed floor first and treat the usage layer as the growth lever, not the primary revenue line, until your customers trust the metering.

The Bottom Line

Usage-based pricing in SaaS went from 27% adoption in 2023 to 38% in 2026, hybrid seat-plus-usage models now cover 61% of the market, and pure per-seat pricing has been squeezed down to just 8%. The companies that made the shift early — Snowflake at 34% revenue growth and 126% net revenue retention, Datadog at 32% growth — are outrunning the roughly 14% median growth rate for public SaaS, while HubSpot's 19% one-day stock drop showed what happens when the market loses confidence in a seat-anchored model. The lesson for founders and investors is the same: price to the value an AI-augmented customer actually captures, not to how many people are logged in.

Compare SaaS growth and multiple benchmarks across pricing models on SaaS Valuations.