What Another SaaS Seat Really Costs (and When Your Own Tool Is Cheaper)

You are about to add the twelfth seat to a tool your team half-likes. The invoice is small. Approving it takes a few seconds, which is roughly what the pricing model was designed to cost you in attention. Nobody in the room asks what that seat will have cost by the time it is retired, because the number on the screen is twelve euros and the number in your head is "fine".

That eleven-second decision, repeated for four years across five tools, is how a small company ends up with a software bill nobody can explain and a process that still needs a spreadsheet to work. This is the arithmetic that should happen before the next seat, and the honest answer about when building your own tool is cheaper and when it very much is not.

The number nobody puts on the invoice

Per-seat pricing is quoted per month because per month is the smallest true number a vendor can show you. The decision you are actually making is multi-year, so that is how it should be priced out. Here are current published list prices, checked this week, extended over five years.

Tool and plan	Published price per seat	10 seats, 5 years	25 seats, 5 years
monday.com Work Management, Standard	€12 / month	€7,200	€18,000
monday.com Work Management, Pro	€19 / month	€11,400	€28,500
Airtable Team (billed annually)	$20 / month	$12,000	$30,000
Airtable Business (billed annually)	$45 / month	$27,000	$67,500

Prices are list, exclude tax, and come from monday.com/pricing and airtable.com/pricing as published in August 2026. Retool, a common middle path for internal tools, splits the bill further: on its Business plan a builder seat is €46 a month and an internal user €14, so the shape of the cost depends on how many people build versus click.

None of these are bad products. The point is only that the five-year column exists whether or not anyone looks at it, and that most SMBs are running three or four of these at once.

Why the bill grows faster than the team

Seat pricing charges you for headcount, not for value received. Hire two people and the bill goes up on every tool they touch, including the ones they will open twice a month. That is tolerable while you are small and quietly punishing once you are not.

Zylo's 2026 SaaS Management Index, published in January 2026, found a median SaaS spend of $9,455 per employee and that organisations leave an average of 36% of their licences unused. Read that carefully before you borrow it: Zylo's data comes from large organisations, over 40 million licences and $75 billion of spend under management, so the absolute figure is not your figure. The direction is what transfers. A meaningful share of what any company pays per seat is paid for access nobody uses.

Two more things push the number up over time and never appear in the first-year plan. The feature you eventually need is usually one tier up, so the per-seat price you signed at is rarely the one you keep. And once a tool holds live operational data, leaving it stops being a pricing decision and becomes a migration project, which is precisely the leverage the vendor is pricing against.

The other column: what a build actually costs

Custom software has the opposite cost shape. It is expensive on day one and cheap on day four hundred. Most of the fear around it comes from imagining an enterprise project, when the thing an SMB usually needs is small: a screen over a real database, a job that runs every night, a form that writes to two systems instead of one person retyping into both.

You can sanity-check the order of magnitude yourself with published rates rather than a brochure. Our own senior white-label band is €35 to €55 an hour. A focused internal tool that takes two developers four weeks is 320 hours, which lands near €14,000 at the middle of that band. That is arithmetic on a published rate, not a quote for your project, and a genuinely simple build comes in well under it. Set against the table above, a €14,000 build is roughly what twenty-five people cost on a mid-tier subscription over four years, and at the end of it you own the code instead of renting access to someone else's.

The honest counterweight is that maintenance is not zero. Software you never touch again is a fantasy sold by people who do not maintain software. Budget for hosting, for a dependency upgrade once a year, and for a few days of changes when your process changes. What you avoid is the per-head escalator: adding your fortieth employee to your own tool costs nothing. If you want the payback math on a specific process rather than a generic one, our automation cost calculator does it in about a minute.

Three times another seat is the right answer

We build custom software for a living and we still talk people out of it most weeks. Buy the seat when any of these is true.

• The tool does the job and you are only adding people to something that already works. Working software at twelve euros a head is not a problem in need of a project.
• The function is a commodity every business does identically. Payroll, accounting ledgers, e-signature, email, video calls. There is no version of your company where building these pays off.
• You are not confident the process will still exist in a year. Renting is the correct way to pay for something you might stop doing, and cancelling a subscription is much cheaper than regretting a build.

Three times the math flips

The signal is never the price. It is a mismatch between what you are paying for and what your team actually does with it.

• The workaround has a name. Everyone knows the spreadsheet, someone owns it, and it exists because the tool cannot handle a step your business depends on. You are now paying a subscription and maintaining a shadow system beside it.
• You are paying full seats for people who look at one screen. Ten warehouse or floor staff who need a status view are ten full licences in most seat models, and a read-only screen over your own data is not a hard thing to build.
• The work happens somewhere the tool cannot reach. Local scanners, printers, a machine on the production floor, a database that is not allowed to leave the building. Cloud-first tools are not being lazy here, they genuinely cannot get to that hardware. We wrote about where that line falls in Custom Automation vs Zapier.

How to test the decision in an afternoon

1. Pull the real usage report from each tool and count how many paid seats logged in during the last month. This is usually the most uncomfortable and most useful number of the day.
2. Write down the workaround in one sentence, including who maintains it and roughly how many hours a week it eats. If you cannot find a workaround, stop here and keep buying seats.
3. Get one fixed quote for replacing the workaround, not for replacing the tool. Scope matters more than vendor here, and a first build should be small enough to be boring.
4. Put both five-year totals on one page, with an honest annual maintenance line on the build side. If the numbers are close, buy the seat. Close is not a reason to start a project.

Do the arithmetic before the renewal

Most companies never run this comparison because renewals are quiet and projects are loud. The tool that costs you the most is rarely the one you complain about; it is the one you stopped noticing.

If you have a process where the software mostly fits and a spreadsheet covers the rest, that is the case worth pricing. Send us the process in three sentences through our contact page and we will tell you in the first call whether a fixed-quote build beats your current seats, or whether you should keep paying for them. We say the second one more often than you would expect.