Seat-based pricing charges a fixed fee for each user (seat) on an account, usually per month or per year, so a 10-person team pays ten times what one person pays. It works when the value a customer gets grows with headcount. It fails when your product lets customers do more with fewer people, which is why many AI products now add usage allowances to the seat.
Key takeaways
- Seats are the easiest model to buy and budget for, and they grow with your customer's team.
- Price the seat on the value one user gets, then add volume tiers, seat types and a minimum.
- If your product reduces headcount, per-seat pricing shrinks your revenue as you succeed.
- The 2026 default is a hybrid: a seat fee plus an included usage allowance.
In the late '90s I worked in IT support.
Every few years a new version of Windows came out, and every business I looked after paid for it the same way: one license per computer. Ten desks, ten licenses. Hire a new receptionist, buy another license.
Nobody questioned it. It felt fair.
When software moved to the cloud, that license quietly became a "seat." Instead of paying a few hundred dollars per PC every few years, you paid a few dollars per person every month. It's the most common way software has been sold for two decades, and it's still the first model most founders reach for.
Then AI arrived and asked an awkward question: what happens to a per-person price when your product means the customer needs fewer people?
Let me show you how seat-based pricing works, where it still shines, and how to tell when it's time to change.
What is seat-based pricing?
Seat-based pricing (also called per-user or per-seat pricing) charges a fixed fee for each person who has access to your product. The bill is simple:
Monthly bill = Number of seats × Price per seat (+ any platform fee)
That's the pure version. In 2026, almost nobody sells the pure version. Here are the variations you'll see on real pricing pages:
| Variation | How it works | 2026 example (US$) |
|---|---|---|
| Classic per seat | Same price for every user | Linear Basic $10, Business $16 per user per month (billed yearly) |
| Active-user billing | Pay only for people who actually use it | Slack's fair billing policy; Pro $7.25 annual / $8.75 monthly |
| Seat types | Different prices for different roles | Figma Professional: Full $16, Dev $12, Collab $3 |
| Platform fee + seats | A base fee for the account, then seats | Devin Teams: $80 a month plus $40 per full developer seat |
| Seat minimum | Must buy at least X seats | Claude Enterprise: $20 per seat plus usage, 20-seat minimum |
| Seat + AI allowance | Each seat includes usage credits | Atlassian includes 25, 70 or 150 Rovo credits per user, pooled across the site |
The last row is where the market is heading. I'll come back to it.
Why did per-user pricing become the default?
Because it's easy for everyone.
For you, revenue is predictable. If 200 customers each have 12 seats at $15, you know you'll bill $36,000 next month. You can forecast, hire and sleep at night.
For your customer, the cost is easy to budget. A manager can work out the cost of adding three people without a spreadsheet or a sales call.
And here's the big one: seats grow with your customer. When your client hires, you get a pay rise. That's land and expand in its simplest form. You win a small team, prove your worth, and the account grows as the team does.
Linear shows this still works. The project-tracking tool sells per seat, reports more than 40,000 paying customers and a net revenue retention of 177% (as reported to ARR Club in August 2026). Its customers keep adding people, so its revenue keeps compounding.
So, seats aren't dead. They're just no longer automatic.
Why is AI breaking the seat?
Picture an accounting practice with 20 staff using a $15-per-seat workflow tool.
That's $300 a month for the software company.
Now the practice adopts AI. Data entry, reconciliations and first drafts get faster. Two years later, the practice serves more clients with 12 staff.
The software is doing more work than ever. But the bill falls to $180 a month. A 40% revenue cut for being useful.
That's the problem in one line: per-seat pricing punishes you when your customer gets more productive.
How WiseTech killed seat fees before AI could kill them first
WiseTech Global's CargoWise runs freight forwarding for some of the biggest logistics firms in the world. In December 2025 it launched CargoWise Value Packs, which remove seat and user fees entirely and charge only on a transactional basis. By February 2026 about 95% of CargoWise customers were on the new model. WiseTech's explanation was blunt: "For SaaS businesses that monetize based on seats or users, AI will disrupt them... our revenue is not diluted by customers becoming more productive." Read the full WiseTech pricing case study.
The data says WiseTech isn't alone. ICONIQ's 2026 survey of about 305 AI builders found consumption pricing rose from 35% to 42% of companies, with businesses blending an average of 1.7 pricing models. And when we looked at the 44 companies that reportedly crossed $100 million ARR in 2026, only four of the 33 whose pricing we could see relied mainly on seats. Eighteen charged mainly on consumption: usage, credits or outcomes. (You can read the whole analysis in how 2026's fastest-growing companies price.)
There's a second problem too. AI features cost real money to run. A light user and a heavy user might cost you ten or a hundred times different amounts in model fees, yet both pay the same $15. That's why GitHub moved every Copilot plan to included credits in June 2026: a quick chat question and a multi-hour coding session were costing users the same amount.
Is seat-based pricing right for my business?
Here's a quick test. Seats are a good fit when most of these are true:
- The value a customer gets grows with the number of people using it (collaboration, communication, shared workspaces).
- Each user costs you roughly the same to serve.
- Your product doesn't replace the people it's priced on.
- Buyers want a predictable bill they can approve once a year.
- You can easily see who has access (and who's sharing a login).
And seats are the wrong fit when:
- Your product does work instead of helping people work (an AI agent, an automation, a report generator).
- One customer's usage can be 50 times another's at the same headcount.
- Customers share logins to dodge the per-seat fee.
- The value lives in something else you can count: projects, transactions, documents, locations, revenue.
If you ticked fewer than three boxes, read the seat vs usage comparison before you go further. Your real value metric might be sitting in that last bullet point.
Value Metric FinderScore candidate value metrics and pick the one that grows with your customers.Open the free tool →How do I set a per-seat price?
Let's build one. Say you sell scheduling software to allied health clinics.
Step 1: Price the value of one user. Ask what one practitioner gets. If your tool saves each physio two hours of admin a week, that's roughly eight hours a month. Even valued at a modest $40 an hour (all illustrative numbers), that's $320 of value. A price of $39 per practitioner per month captures about an eighth of it, which leaves the customer an easy yes.
Step 2: Add a platform fee or a minimum. Tiny accounts can cost as much to support as big ones. A small base fee, or a minimum of two seats, covers onboarding.
Step 3: Use seat types. Practitioners get full seats. Front-desk staff who only book appointments get a cheap or free "reception" seat. That keeps the price fair and gets more people inside the product.
Step 4: Add volume bands. Reward growing clinics, but keep the bands visible:
| Practitioner seats | Price per seat per month | Example bill |
|---|---|---|
| 1–5 | $39 | 4 seats = $156 |
| 6–20 | $34 | 12 seats = $408 |
| 21–50 | $29 | 30 seats = $870 |
| 51+ | Talk to us | Custom |
| Reception seats | Free (up to 5), then $5 | 3 = $0 |
Illustrative prices for a hypothetical product.
Step 5: Offer annual billing. Most seat products discount annual plans by around 15% to 20%. Work out what you can afford with the annual vs monthly guide.
Step 6: Include an allowance if you have AI costs. If your product runs AI, include a monthly allowance per seat (say, 200 AI summaries) and sell top-ups. That protects your margin from your heaviest users.
When you've got your numbers, lay them out as a good-better-best table. The Pricing Table Builder does that for you in a few minutes.
Is the seat still right for you?
Answer a few questions and the Pricing Model Picker tells you whether seats, usage, credits or a hybrid fit your business.
How do I fix seat pricing without blowing it up?
Maybe you already sell per seat and you can feel the cracks. You don't have to tear it all down. Here are five fixes, from gentlest to boldest.
1. Make light users free. MaintainX gives requesters free accounts, Payhawk includes unlimited employee seats and charges for admins and volume, and Linear's free plan allows unlimited members. You charge the people who do the work, and everyone else becomes a reason not to leave.
2. Bill active users only. Slack's fair billing approach removes the "we're paying for people who left" complaint that causes downgrades at renewal.
3. Bundle AI into the seat and lift the price. Slack folded AI into Business+ in June 2025 and raised it from $12.50 to $15 per user. Notion retired its separate AI add-on and bundled AI into Business. This works when the new capability is real. Otherwise it feels like a tax. (See how Slack put AI in the box.)
4. Add a usage allowance. Keep the seat, include a pooled allowance of credits, and charge for extra. This is the hybrid pricing model, and it's now the most common structure for AI products sold to teams.
5. Change the meter. Move from seats to the work itself, as WiseTech did with transactions and Procore did with construction volume. It's the biggest change and needs the most care: migrate customers with a better bundle, give notice, and explain it as alignment ("you pay when you get value").
My take
I wouldn't rip seats out of a product where collaboration is the value. Seats are a great way to grow with a customer. But if you're building anything that replaces work rather than supports it, don't launch on seats. You'll be pricing yourself on the one number your product is designed to shrink.
What to do next
Write down three numbers for your business: how many users your average customer has today, how many they'll have in two years if your product works brilliantly, and what else grows when they get value (projects, clients, transactions).
If the user number goes down while the others go up, seats are the wrong meter. Start with the Pricing Model Picker, then read about credit-based pricing and usage-based pricing to see what a move could look like.
If the user number goes up, you're in good shape. Tune your bands, free up your light users, and keep growing with your customers.
Frequently asked questions
Is per-user pricing the same as per-seat pricing?
Mostly, yes. Both charge for each person with access. Some companies use 'seat' for a license that can be reassigned and 'active user' for someone who actually logs in during the billing period. Slack, for example, bills only for active members under its fair billing policy, which removes the fear of paying for people who never use the product.
Should I charge for admin or viewer accounts?
Usually not at full price. Charge for the people who get the core value (the 'doers') and make light users cheap or free. Figma sells Full, Dev and Collab seats at different prices, and MaintainX gives requesters free accounts. Free light users spread your product through the customer's business and make the paid seats stickier.
What is a seat minimum?
A seat minimum is the smallest number of seats a customer must buy on a plan. Anthropic's Claude Enterprise plan, for example, has a 20-seat minimum. Minimums protect you from enterprise-level support costs on tiny accounts. Use them on your top tier, not your entry tier.
How much discount should I give for more seats?
Enough to reward growth without collapsing your average price. A common shape is full price for the first band of seats, then a step down of around 10% to 20% per band. Make the bands visible on your pricing page so a team lead can work out the cost of adding people without calling you.
Does seat-based pricing work for small businesses outside software?
It can. A bookkeeping practice can price a monthly plan per staff member on payroll, a gym corporate plan can price per employee, and a training business can price per learner. The test is the same as for software: does the customer get more value as they add people? If yes, a per-person price is easy to explain.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- WiseTech Global 1H26 results briefing scripts (Feb 2026)
- Slack: June 2025 pricing and packaging announcement
- Slack pricing page (checked 27 Oct 2026)
- Figma pricing page (checked 27 Oct 2026)
- Claude pricing (checked 27 Oct 2026)
- SaaStr summary of ICONIQ's 2026 State of AI report (10 Jul 2026)
- ARR Club: Linear hits 40k paying customers with 177% NRR
- Atlassian Support: Rovo usage limits
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.