Modern Pricing & UBB
February 26, 2026 ubb

Seat-Based Pricing in 2026 Is a Coping Mechanism

Seat-based pricing in 2026 is like still using Internet Explorer — technically it works, everyone knows it's wrong, and somehow the enterprise still insists on it. We know better. The data is in. OpenView found that 61% of SaaS companies have adopted some form of usage-based pricing. Snowflake, Stripe, Twilio, AWS — the biggest names in infrastructure are all printing money on consumption models. And yet here we are: still counting butts in seats.

Why It's Kind of Dumb

The logic of seat-based pricing made sense once. You built software for humans. Humans had desks. Desks had licenses. Finance understood it. Procurement budgeted for it. Simple. But that was before a single AI agent could do the work of 50 Zendesk seats. Before Snowflake proved you could grow NRR to 150%+ just by charging for compute consumed. Before Stripe built a trillion-dollar payments network by taking a cut of every transaction — not charging per API key holder.

When Twilio launched in 2008, it didn't say "we'll charge $500/month per developer with API access." It said: $0.0085 per SMS. Pay for what you use. That model let a two-person startup access enterprise-grade communications infrastructure on day one — and pay nothing until they had actual customers. That's why developers loved it. That's why it scaled. Charging per seat would have killed it before it started. AWS didn't become a $100 billion business by making people buy EC2 licenses. They charged for the compute you actually ran.

Why Companies Keep Doing It Anyway

So why is seat-based pricing still the default for most B2B SaaS in 2026? Honestly? It's not entirely stupid. Three very human reasons:

1. Finance departments speak fluent "per seat." Tell a CFO your software costs $45/user/month and they can forecast it in their sleep. Tell them it's usage-based and suddenly someone needs to build a model, find a spreadsheet, and schedule three meetings about projected consumption. Procurement loves certainty. Annual contracts close faster when the number on the PO doesn't change based on how many API calls your engineers made in November.

2. Seat-based revenue is predictable. Usage-based pricing is great until a recession hits and your customers quietly throttle their workloads. Snowflake found this out the hard way in 2023 — they had to cut guidance because customers optimized their query patterns. That doesn't happen with a locked-in annual license. Predictable revenue isn't just a finance preference; it's how you raise your Series B without sweating through your shirt.

3. Usage infrastructure is genuinely hard to build. Real-time metering, usage dashboards, overage alerts, bill shock prevention — this is complex plumbing. Until Orb, Metronome, and Lago existed, most engineering teams couldn't justify the investment. Seat billing is a Stripe webhook and a spreadsheet. Usage billing is an entirely new data pipeline. When you're trying to ship features, "fix pricing infrastructure" rarely wins the sprint planning vote.

The Part Where We Do the Math

But "not entirely stupid" is not the same as right. UBP companies show 54% more revenue scale than traditional SaaS peers at IPO. Top-quartile NRR for usage-based companies hits 120% — 10 full points above pure subscription companies. The math is not subtle. More importantly, the world is changing faster than your pricing model. Every AI agent you deploy means one fewer human seat. Your product gets better, your customers use it more, and under seat pricing… your revenue stays flat. That's not a business model. That's a charity.

The SaaS companies still 100% on per-seat in 2026 aren't dumb — they're comfortable. And comfortable is fine. Right up until a usage-based competitor shows up with a lower entry price, better expansion economics, and customers who actually feel good about their bills.

Seat pricing isn't dead. But it's definitely on a diet.

Sources

← Seats & Per-User · All posts