Per-seat SaaS pricing had a good run. For the better part of two decades, it was the default way enterprise software got sold - pick a plan, count your users, pay monthly. It was simple, predictable, and easy to budget. Vendors loved it because revenue scaled with headcount, and headcount mostly went up.
Though, that relationship between headcount and software value is quietly breaking down.. and 2026 is the year it's becoming impossible to ignore.
Why the Seat Model Worked in the First Place
The logic behind per-seat pricing was always a proxy. You weren't really paying for outcomes, but you were paying for access. More people using the tool meant more value extracted from it, so seat count was a reasonable stand-in for that value. Hire 10 more salespeople, buy 10 more CRM seats. The math held up.
That changed when AI agents entered the picture. An AI agent doesn't log in with credentials, nor show up in an admin's license dashboard. It just executes work - sometimes hundreds of tasks - without ever occupying a seat in the traditional sense. When one agent handles what previously required a team of 10, per-seat pricing becomes disconnected from the value being delivered.
As Satya Nadella framed it earlier this year, seats are increasingly becoming "just entitlement to some consumption." Deloitte's 2026 TMT Predictions put it plainly - the pricing unit no longer maps to the work being done.
The Numbers Behind the Shift
Seat-based pricing fell from 21% to 15% of companies in just 12 months, while hybrid models surged from 27% to 41%, according to Growth Unhinged's 2025 State of B2B Monetization. IDC predicts that 70% of software vendors will have refactored pricing away from pure per-seat models by 2028.
And the credit model - you know, where you buy a bucket of usage rather than a fixed number of seats - was the defining pricing innovation of 2025. Adoption of credit based models jumped 126% year over year across the top 500 SaaS companies, according to the PricingSaaS 500 Index.
The Four Models You'll See in 2026
There's no single replacement for the seat model. The industry is landing in different places depending on the product and the customer. Here's a breakdown -
| Model | How It Works | Best Fit | Watch Out For.. |
|---|---|---|---|
| Per-Seat | Fixed fee per named user | Tools where every user has similar, regular usage | Overcharging when AI reduces how many humans need access |
| Usage Based (Credits / Tokens) | Pay for what you consume. API calls, tokens, actions | AI tools, developer tooling, anything with variable usage | Unpredictable bills; heavy users can blow budgets fast |
| Hybrid | Base subscription fee + variable usage on top | Most teams right now - balances predictability with scale | Complexity; need to watch both the floor and the meter |
| Outcome-Based | Pay per measurable result (tickets resolved, leads qualified) | AI support, sales tools, anything with a clear output | Requires agreed definitions of 'success' before you sign |
43% of SaaS companies now use hybrid models as their primary structure, with adoption expected to hit 61% by the end of 2026, per Chargebee's 2025 State of Subscriptions Report. Hybrid has become the most common model because neither vendors nor buyers are fully ready for pure outcome based pricing yet, and nobody wants to go back to pure per-seat either.
The Platforms That Have Already Moved
The most visible example right now is GitHub Copilot. As of June 1, 2026, GitHub officially switched all plans to usage based billing, replacing its flat premium request model with token based AI Credits. Every interaction now costs credits based on token consumption - input, output, and cached context. Base plan pricing stayed the same ($10/month Pro, $19/user/month Business, $39/user/month Enterprise), but the included credit pools are finite, and heavy agentic workflows burn through them fast. Some developers reported significant cost jumps almost immediately.
GitHub explained the move directly - Copilot is no longer the same product it was a year ago. Agentic workflows consume far more compute than a simple autocomplete suggestion, and flat-rate pricing couldn't reflect that sustainably.
Other platforms moved earlier and more quietly:
- Zendesk launched outcome based pricing for AI agents in August 2024. $1.50 per committed automated resolution, $2.00 pay as you go. You pay when a ticket is resolved by AI without human intervention.
- Intercom Fin charges $0.99 per resolved conversation. The model earns revenue when the product works.
- Salesforce Agentforce launched at $2 per AI conversation, then expanded to multiple billing options after enterprise pushback - including $125/month for unlimited internal use and pay as you go consumption at $0.10 per action.
- Figma, HubSpot, Airtable, and Monday.com all adopted credit based models in 2025.
Most of these platforms didn't abandon per-seat entirely. They layered usage or outcome components on top of a base subscription. That's the hybrid approach in practice, and it's the path most teams are landing on.
Rethinking Your SaaS Stack?
If your team is paying per-seat for tools that mostly run automated workflows, it might be worth looking at what a purpose-built integration could save you.
If You're Buying SaaS - What to Check at Your Next Renewal
The shift is already affecting renewal conversations, whether you realize it or not. Tropic coined the term "AI Tax" to describe the 20–37% price uplift vendors are slipping in at renewal by bundling AI features into existing tiers or migrating customers to more expensive AI-inclusive SKUs. You might not see a seat count change, but you'll see the bill go up.
A few things to look at before you sign:
- How does the tool measure value?
- If your team measures value in outcomes - tickets closed, reports generated, decisions made - then paying for seats that don't reflect that is already a misalignment.
- What happens when your AI usage spikes?
- Credits and token based models can create unpredictable bills. Ask for a usage forecast before committing to any plan that meters AI features separately.
- Is an outcome agreement defined?
- For outcome based pricing, get the definition of 'success' in the contract before signing. Zendesk requires this explicitly, and most buyers miss it.
- Are you paying for access you no longer need?
- If AI has automated work that previously required human users, audit whether those seats are still earning their keep.
If You're Building Software, This Is Architecture
For teams building products, the pricing shift has a less obvious but equally real implication - the value metric you choose to build around is now a strategic decision.
If you're building on top of third party AI services, your own cost structure is usage-based whether you like it or not. Running a frontier model costs real money per token, and a flat rate subscription on top of that creates a margin problem the moment your power users show up. Credit based models are increasingly how product teams are solving this; they let you pass through cost variability to customers without surprising them with unpredictable bills.
For teams building internal tools or replacing SaaS subscriptions with custom built solutions, this moment is actually worth paying attention to. If your team is paying per-seat for a tool that mostly runs automated workflows, you're paying for access to something that doesn't need the access. A purpose built integration or internal tool (one you own and control) doesn't come with a seat count, a renewal negotiation, or an AI Tax.
That's not always the right call. Off the shelf SaaS still makes sense for a lot of use cases. But the calculation is worth running, especially as vendor pricing becomes harder to predict. If you're evaluating that tradeoff, our custom API and integration services are one place you can start.
TL;DR
Per-seat pricing isn't going away completely. Most teams are going to land in hybrid territory for the next few years, paying a base fee with variable usage on top. But the era of flat-rate, headcount-scaled SaaS is winding down, and 2026 is when the major platforms are putting that on paper.
If you're buying software, the renewal conversation is the moment to push back and ask whether the pricing model still reflects how your team works. If you're building software, the choice of value metric is a design decision that touches revenue, margins, and customer retention all at once. Worth getting right before it's baked in.
