Build or buy? How to tell which one you actually need

Buy, unless your seat count says otherwise. The operations app priced below costs $48,000 – $73,000 to build, over 15–23 weeks. Five years of monday.com Work Management Pro for ten people is $11,400. The math doesn’t turn over until roughly 53 seats, and most businesses asking this are a long way short of that.

$48,000 – $73,000: Building the operations app described below
53 seats: Where five years of monday.com Work Management Pro matches the build price
77 seats: The same count once the build carries five years of its own changes

Written by

WhatWillMyAppCost

We build the estimator this site runs on, and we design and build custom mobile and web software for a living. Every figure below is the estimator’s own output, except the published vendor prices, which are read off each company’s own page and dated.

You pay for four subscriptions. None of them quite fits, so every Tuesday somebody spends the morning copying rows out of one and pasting them into another. Your operations manager has started saying you should build your own. Your accountant has started asking what that would mean.

What a real build of that tool costs, the seat count where five years of subscription catches it, what the comparison misses on both sides, and the four situations that favor building, three of which have nothing to do with headcount.

Should you build or buy?

Buy. For almost everyone asking, the product on the shelf is cheaper for years, and somebody else is awake at 2am when it breaks. Four situations flip that, and the first is arithmetic: your headcount outgrows per-seat pricing. The other three are further down.

Both halves of that need a number, and only one of them is easy to find. You already know what you pay per seat. The build figure is below, along with the seat count where the two meet. If your current system is a shared spreadsheet instead of a subscription, the arithmetic works differently. And if you don’t yet have a description to price, start with the brief.

$48,000 – $73,000: The build
$28,500: Five years of monday.com Work Management Pro for 25 people
$67,500: Five years of Airtable Business for 25 people

What does a custom build cost?

$48,000 – $73,000 over 15–23 weeks for the tool in that description: sign-in, three kinds of user, a web side for the office and phones for the crew, documents filed against a job, search, a few reports, a connection to your accounting system, and everything sitting in the current tools moved across.

$48,000 – $73,000: Everything in that description, 15–23 weeks
−31%: What a narrower first version takes off, about $18,500

The range is wide because the pieces are. The largest line is the office portal at about $11,000, and the office half of a build is reliably the half nobody budgets for. Giving three groups of people three different sets of screens is another $8,500. Reports run $3,500, the accounting connection $2,000, and moving the old data across $2,000. How the estimate is put together covers the rest.

A narrower first version takes about $18,500 off, 31% below the full build, mostly by shrinking the office side and holding the reports back. That’s a real option, and what a first version should cost goes into what survives the cut.

How many seats before building is cheaper?

Roughly 53 against monday.com Work Management Pro, and roughly 22 against Airtable Business. Both are five years of subscription at the annual rate, set against the middle of the build range. Treat those as the floor of the seat count, not the number: they compare five years of subscription against a build price that stops the day it ships. Put five years of changes on the build side too, at the rate this page prices them, and the two counts move to about 77 and 33.

monday.com Work Management Pro lists at $19 per seat a month on annual billing as of August 2026. Call it $228 a seat a year. Ten people over five years is $11,400. Twenty-five is $28,500. Fifty is $57,000, which lands inside the build range without quite reaching the middle of it.

Airtable Business lists at $45 per user a month on annual billing, checked the same day. Ten people over five years is $27,000. Twenty-five is $67,500, which has already gone past the middle of the build range. The tier you land on moves this answer as much as the headcount does.

How the comparison gets described: A subscription is a few hundred a month and a build is fifty grand. Put that way it isn’t a decision, it’s a sanity check.
How it reads at fifty seats: $57,000 of monday.com Work Management Pro against a $48,000 – $73,000 build, over the same five years. Now it’s a real question, and headcount is the variable that got you there.

Ten people on Airtable Business for five years is $27,000. The build is $48,000 – $73,000. At ten seats this isn’t close, and no amount of dissatisfaction with the product changes that. Count your seats before you price anything else.

What does the five-year comparison leave out?

Money on both sides, and not the same kind of money.

On the build side: Hosting, and every change after launch. A change gets priced the way the original feature was: adding reports to the build above runs about $3,500, a second system connection about $2,000. Assume a few of those a year and the five-year figure moves.
On the subscription side: Seats. The price is per person, and headcount goes up more often than it comes down. Modules get added. Annual rates get revised. A figure that beats a build at this year’s headcount may not beat it at next year’s.
On both sides: The move itself. Getting your data out of one system and into another costs money whichever way you go. In the build above it’s about $2,000, and switching between two products isn’t free either.

This site prices builds, not upkeep. Read every build figure here as what it costs to get to a working thing, with hosting and changes on top of it.

When is buying obviously right?

When your problem is an ordinary version of a common problem. Most problems are, and there’s no prize for having an unusual one.

What people say: “Nothing off the shelf does what we do.”
What’s usually underneath it: Something does most of it, and the missing part is a habit the business could change more cheaply than it could build software. Worth testing before it gets priced, because it’s the difference between a subscription and five figures.

Under twenty people will use it, and that number isn’t about to double.
A product exists that does most of the job, and the gaps annoy people without stopping them.
You need it working this quarter.
Nobody on your side wants to own a piece of software for the next five years.

When is building the right call?

Four, and only the first one is arithmetic. Outside them, buy.

Seat count. Past roughly 53 people on a $19 seat, or roughly 22 on a $45 one, five years of subscription costs more than the build did.
No product models the way you actually work. Not “the product is clunky”. The specific step that wins you business, which nobody built a product around because nobody else does it that way.
You’re paying people to move data between tools. Two people at half a day a week is real money, it compounds, and you can measure it before you commit to anything.
The product is going away, or its price just moved somewhere you can’t plan around.

The first and third you can settle this week with a payroll report and a stack of invoices. The second is the judgment call, and it’s the one worth arguing about internally before anybody outside gets asked for a number.

Can you buy most of it and build the rest?

Often, and it’s the cheapest way to get the custom part by a wide margin. Keep the product for the ordinary work, build the one piece it can’t do, and connect them. The accounting connection in the build above runs about $2,000, which is what a small piece of custom software bolted onto something you already pay for tends to look like.

The catch is that you keep paying for the seats and you’re still living inside somebody else’s product decisions. That’s a good trade when the custom piece is small and the product is good. It’s a bad one when the custom piece is most of the job, because then you’ve paid for both.

The operations app above takes $48,000 – $73,000 and 15–23 weeks to build. Five years of monday.com Work Management Pro matches it at roughly 53 seats, Airtable Business at roughly 22.

Below those numbers, buy, and spend the difference on the part of the business that earns. Above them, or when no product models the step your business wins on, a build starts to pay for itself. Count the seats before anything else, because most of the answer follows from that one number.

FAQ

Questions about building versus buying

Is it cheaper to build custom software or buy off the shelf?

Buying, until your seat count gets large. The operations app priced on this page costs $48,000 – $73,000 to build. Five years of monday.com Work Management Pro at $19 a seat comes to the same money at roughly 53 seats, and Airtable Business at $45 a seat at roughly 22. Both rates are annual billing, checked August 2026.

How do you work out the total cost of ownership of custom software?

Start with the build, then add hosting and the changes. Changes get priced like features: in the build above, adding reports is about $3,500 and a second system connection about $2,000. A build figure on its own covers getting to a working thing, and says nothing about keeping one.

At what point does per-seat software get too expensive?

When five years of it passes what the replacement would cost to build. Multiply your per-seat rate by twelve, by your headcount, by five. Set that against a build estimate. On the figures here that crossover lands at roughly 53 seats on a $19 seat and roughly 22 on a $45 one.

What does a custom software budget usually miss?

Hosting, changes, and your own people’s time. Somebody at your company has to answer questions during the build and own the thing afterward. Moving your existing data across is a line item too, about $2,000 in the build above, and it’s the one most often left out of the comparison entirely.

Can custom software be cheaper than SaaS in the long run?

Yes, above a certain headcount, because a build is a fixed cost and a subscription is a per-person one. Fifty people on a $19 seat is $57,000 over five years. Ten people on the same seat is $11,400. The build price barely moves between those two, which is the whole reason the crossover exists.

How long does custom business software take to build?

15–23 weeks for the tool described here, which covers three kinds of user, a web side, phone access, search, reports, an accounting connection and a data move. A narrower first version lands in 13–20 weeks. A subscription is running by Friday, and that difference matters more to some businesses than the money does.

What happens to custom software if the people who built it move on?

It keeps running, and the next team charges more for their first month than the last team charged for their third. That risk is real and it is manageable: it comes down to how much of the setup is written down and how ordinary the technology choices were. Both are fair things to talk about while a build is being scoped, not after.

Should a small business build its own software?

Usually not. At ten seats the arithmetic is not close: ten people on Airtable Business for five years is $27,000, against $48,000 – $73,000 to build. It tightens faster than people expect, though. Twenty seats on the same plan is $54,000, which is already inside the build range. The exception is a small business whose margin depends on doing one thing differently from everyone else, because that is precisely the thing no product will ever cover.