Monday.com lists Work Management Pro at $19 per seat per month (monday.com/pricing, checked 5 September 2026). Four hundred seats is $7,600 a month and $91,200 a year. That number is correct, and it is the least interesting number in this article, because it is the one the vendor expects you to negotiate against. The interesting numbers are the ones that do not appear on the invoice.
Start with three years, not one
Nobody signs a work management contract for a year. You sign for a year and then renew for as long as the boards exist, which is as long as the company exists. So the honest unit of comparison is three years, and three years of a renewing contract is not three times the first year.
Assume a 7% uplift at each renewal, which is a mild assumption by current standards. Year one is $91,200. Year two is $97,584. Year three is $104,415. Three years is $293,199, and at the end of it you own a login.
If your renewal letter has ever contained a larger number than 7%, use that instead. The arithmetic below only gets more one-sided.
The seat count is not the seat count
Open the invoice, not the admin console. The quantity on the invoice line is what you pay for; the number of humans in the admin console is a different number, and in every audit we have run they disagree.
Two mechanics do most of the damage. First, seats are sold in blocks, so a company with 412 people does not buy 412 seats. Second, deactivation is not always the same as de-provisioning: find out, in writing, whether removing a user mid-term returns the seat to the pool or whether you carry it to the renewal boundary. Ask your account manager that question by email and keep the reply.
Then run one more count: how many of those paid seats logged in at all in the last ninety days. The audit log will tell you. This is the single cheapest piece of work in the whole exercise and it routinely finds double-digit percentages of paid seats that nobody has touched since onboarding.
The viewer-seat trap
Here is the mechanism that makes per-seat work management more expensive than it looks. The people who most need to see the plan are the people who will never edit it: the finance business partner, the account director, the ops manager, the person from legal who needs to know when the thing ships. Under per-seat pricing, letting someone look costs exactly what letting someone work costs.
So organisations do the rational thing and do not give them access. And then someone rebuilds the plan by hand, every week, in a slide. The cost of the viewer-seat trap does not appear on the invoice; it appears as four hours a week of a delivery manager exporting to a spreadsheet, and as a business that makes decisions from a screenshot that was true on Tuesday.
Monday.com does have an answer to this: free viewer seats appear on the Enterprise plan, which is quoted by sales rather than published. That is worth reading twice. The escape from the viewer tax is itself an upgrade conversation, and its price is whatever the account team thinks you will pay.
A database row does not have a price. Once the system runs in your own account, the question "should this person be able to see the plan?" stops being a procurement question and goes back to being an access-control question, which is where it belongs.
What else lands on the line item
- Automation and integration actions are metered per plan. Find your monthly cap and your current burn. When you exceed it, the remedy offered is a plan upgrade, not a bill for the overage.
- The administrator. Someone maintains the boards, the automations, the permission scheme and the integration credentials. In a 400-seat estate that is rarely less than a quarter of a person, and it does not go away when you replace the tool — but you should count it on both sides of the comparison.
- The annual security review, the DPA renegotiation, and the vendor questionnaire, on a system that now holds every estimate, slip and dependency in the business.
- The exit cost. A CSV export of your boards preserves rows. It does not preserve automations, mirror columns, formula columns, update threads, file attachments or permissions. The longer you stay, the larger the part of the system that only exists inside the vendor.
What the replacement costs
Cadence is our reference implementation for this category. At the reference size — 400 seats, roughly two million items, 25 GB of attachments, production and non-production, single region — the infrastructure runs at about $780 a month on AWS and $830 a month on Azure. That is $9,360 a year, against a three-year compute savings plan, with Aurora at a 0.5 ACU floor and non-production stopped outside business hours. It excludes support plans and egress above a terabyte a month.
The build is eight weeks with two engineers to parity on the ten core capabilities, including migration of the existing boards, items, updates and files. Eight weeks of two engineers is about 3.7 engineer-months. At $22,000 per engineer-month — the figure our calculator defaults to, and the one you should replace with your own — that is roughly $81,000.
Then add the part most build-versus-buy models leave out: someone has to run it. Budget a quarter of a platform engineer, priced at your own internal rate. If you are not willing to fund that in year three, stop reading and renew the licence — that is a real answer, and it is the fourth of the four cases where we tell clients not to do this.
The arithmetic, in full
Rented, 3 years $91,200 + $97,584 + $104,415 = $293,199 Owned, 3 years $81,000 build + (3 x $9,360 infra) = $109,080 Net over 3 years = $184,119 Annual saving $91,200 - $9,360 = $81,840 Payback $81,000 / $81,840 = ~12 months Excludes, on both sides: the internal operator, your internal project time, and any migration weekend you would rather not think about. Add them and payback lands closer to 14 months.
Fourteen months is the honest headline. It is not a dramatic number and we would rather publish it than a dramatic one, because the dramatic ones are the reason nobody trusts build-versus-buy models.
When not to bother
Run the same model at 150 seats. Monday.com Pro at 150 seats is $34,200 a year. The infrastructure does not fall proportionally — a database is a database — so call it $5,500 a year rather than $9,360. The saving is roughly $28,700 and the build has barely shrunk, because the ten core capabilities are the same ten capabilities. Payback approaches three years, before the operator, and three years is longer than most sponsors last in the role.
Below roughly 150 seats, per-seat pricing is genuinely good value and we will tell you so. There are three other cases where the answer is also no: when the thing you are actually buying is a vendor attestation you can hand to an auditor; when the value is a network you cannot rebuild, such as a supplier directory or a marketplace of two hundred integrations; and when there is no team and no budget line to own the result in year three.
If the honest answer is to keep the licence, that is what the report will say. We would rather lose the work in week one than lose it in month five.
What to do this week
- Pull the last invoice and write down the quantity on the line, not the headcount in the console.
- Export ninety days of audit log and count distinct users who logged in, then count distinct users who did anything other than open a board view.
- Ask your account manager, by email, what the renewal uplift will be and whether mid-term deactivation returns a seat.
- List the automations, mirrors and formula columns that would not survive a CSV export. That list is your real switching cost.
- Run your own numbers through the calculator on this site and be suspicious of any payback under nine months.
None of that takes an engagement, and all of it is worth knowing whether or not you ever replace anything. The point is not that Monday.com is bad software. It is good software with a pricing model that charges you for visibility, and visibility is the thing you most want to give away inside your own company.
Prices quoted here are vendor list prices with the date we checked them. If a figure has gone stale, tell us and we will correct it — support@conseiltek.com.