Zendesk Pricing Explained: The Two Meters on Your Bill
People rarely get caught out by the price on the pricing page. They get caught out by the shape of the bill: one meter counting people, another counting how well the AI is doing. Here is how those two fit together, and how to work out your own real number.
A note on numbers
We do not quote Zendesk list prices here. Published amounts change, discounts are negotiated, and a stale figure in an article is worse than no figure. What does not change is the structure, and the structure is where the surprises live. Every EzyConn price on this page is ours and is exact.
There are two meters, not one
Meter one counts people. Zendesk charges per agent per month. Per-agent pricing means every person on the rota costs the same, even a weekend stand-in who logs in twice. It is not a usage charge. It is a licence for a chair, and the chair costs the same whether someone is sitting in it or not.
Meter two counts automated resolutions. An automated resolution is a conversation the AI finished by itself, with nobody stepping in. These are billed separately from the seats, and Zendesk does not publish a single public per-resolution figure, so you have to get it from your rep and get it in writing.
Two meters is the whole story. Everything else, the tiers, the add-ons, the annual commitment, is decoration on top of those two lines.
The formula
monthly bill = (agents × tier price) + (automated resolutions × resolution rate) + add-ons
Three of those four numbers are things you can find out today by opening an invoice and counting heads. The fourth, the resolution rate, is the one most teams never write down, and it is the one that decides whether a busy month costs the same as a quiet one. If you take nothing else from this article, take that: go and find that number.
What happens at 5, 10 and 25 agents
We cannot put a Zendesk figure in this table without inventing one, so the middle column describes what the arithmetic does to you instead. The right-hand column is our own pricing, which is exact.
| Team size | On per-agent pricing | On EzyConn |
|---|---|---|
| 5 agents | Five seat fees every month, even if two are part-time | $95/mo on Professional, with room for 10 agents |
| 10 agents | Ten seat fees, plus the AI meter running underneath | $95/mo on Professional, 10 agent seats plus the owner |
| 25 agents | Twenty-five seat fees. The headcount is the bill. | Enterprise: talk to us, priced on seat volume |
The point is the slope, not the starting point. On a per-agent model the line goes up every time you add a person, so growing the team is a budget conversation. On a per-workspace model the line moves in a few known steps, when your volume or your team outgrows a plan, so hiring inside a plan costs nothing extra. Neither is morally superior. They just fail in different directions.
Work out your real number in ten minutes
- Count every login, not every full-time employee. Include the part-timer, the weekend cover and the manager who checks in once a month.
- Find your per-agent rate on the actual invoice, not the website. Negotiated deals differ.
- Multiply. That is your floor, the amount you pay in a month with zero customers.
- Ask your rep, in writing, for the per-resolution rate and whether any allowance is included before metering starts.
- Multiply that by last month's automated resolutions.
- Add every add-on line on the invoice. Read all of them. There are usually more than you remember.
- Divide the total by the number of conversations you handled. That is your true cost per conversation, and it is the only number worth comparing between tools.
Three things that move the bill
Hiring and the rota. Seasonal cover, maternity cover, a night shift, an intern. Each is a full seat. Teams who plan around this end up understaffing the evening rather than paying for it.
Success. The better your help centre and your AI get, the more resolutions you buy. Most software gets cheaper per unit as you use it more. This gets more expensive in absolute terms exactly when it is working best, which is a strange thing to explain to a finance team.
Tier jumps for one feature. The classic trap. You need a single capability, it lives one tier up, and the upgrade applies to every seat you own. A $10 difference in tier price on 25 agents is $250 a month for one checkbox.
The seat-sharing trap
When seats are expensive, teams share one login for the weekend or the overflow person. It saves money and it wrecks three things at once. Your reporting can no longer tell you who answered what, so performance conversations become guesswork. Your audit trail stops being an audit trail, which matters if you are ever asked to prove who accessed a customer record. And handovers get messy because two people are typing as the same identity. This is not a Zendesk flaw. It is what any per-seat model does to a small team under budget pressure.
What flat pricing looks like instead
For contrast, here is the whole EzyConn price list. Prices are per workspace, not per agent, there is no per-resolution fee, AI runs on GPT-4o and Claude and is included on every plan including the free one, and annual billing takes 20% off. What you buy is a set of seats and a monthly message allowance.
| Plan | Price | Agent seats | Messages/mo |
|---|---|---|---|
| Free | $0 forever | 1 plus the owner | 500 |
| Starter | $25/mo | 3 plus the owner | 5,000 |
| Professional | $95/mo | 10 plus the owner | 25,000 |
| Enterprise | Talk to us | Set by seat volume | Agreed with you |
The free plan is free forever rather than a trial, and includes 1 agent seat plus the owner, 500 messages a month and one chatbot. Details are on the pricing page, and the direct comparison lives on the Zendesk alternative page.
Which model is cheaper for you
Honestly, it depends on the ratio between your headcount and your volume. Per-agent pricing suits deep, low-volume, high-value work.
Flat pricing suits the opposite shape: several people, each doing support alongside another job, handling a steady stream of fairly repetitive questions. That is most small teams. If you want the mechanics of the other billing models in this market, we compared them in AI support pricing models compared.
Common questions
How much does Zendesk cost?
There is no single answer, because the bill has two moving parts. You pay a per-agent subscription for every person with a login, set by which tier you are on, and you pay separately for automated resolutions handled by the AI. Your monthly total is agents multiplied by tier price, plus resolutions multiplied by the resolution rate, plus any add-ons. Ask your rep for both rates in writing before you compare anything.
What does per-agent pricing actually mean?
Every person who needs a login costs the same amount every month, regardless of how much work they do. Your busiest full-timer and a weekend stand-in who logs in twice a month are billed identically. It also means the cost of covering evenings, holidays or a seasonal rush is the same as hiring full-time, which is why so many teams end up quietly sharing one login.
What is an automated resolution and why is it billed separately?
An automated resolution is a conversation the AI finished on its own, without a human stepping in. Zendesk meters those apart from the per-agent subscription, so the AI has its own line on the invoice. The awkward consequence is that the better the automation works, the more of them you buy. A quiet month is cheap and a successful month is not.
Is Zendesk pricing per agent or per ticket?
The subscription is per agent, not per ticket. Volume does not raise the seat cost directly. The separate meter is on automated resolutions, so ticket volume reaches your invoice only through the share of it the AI handles. That is the opposite of per-workspace pricing, where a plan includes a set number of seats and messages and the price stays put inside it.
How does EzyConn pricing compare with Zendesk?
EzyConn charges a flat monthly amount per workspace, not per agent, with no per-resolution fee. Free is $0 forever with 1 agent seat plus the owner and 500 messages a month, then paid plans are Starter at $25 a month with 3 agent seats and 5,000 messages, and Professional at $95 a month with 10 agent seats and 25,000 messages, with 20% off on annual billing. AI on GPT-4o and Claude is included on every plan. You buy a plan sized to your team and volume rather than paying for each person.
One line on the invoice
One price per workspace, no resolution fees, AI included on every plan. Start on the free plan and see what your real conversation volume is before you pay anyone anything.
Start freeEzyConn competes with Zendesk, so treat this as an informed but interested account. We describe the Zendesk billing model rather than quoting figures, because published amounts change and negotiated deals differ. Always check your own invoice.