A decision, not a preference

Build or buy,
decided honestly

Almost every article on this question is written by someone who sells one of the two answers, which is why they all conclude so confidently. The honest version is that buying is right most of the time, and the exceptions are specific enough to name. Four questions settle it, and none of them is about the licence price.

Question one: is this how you are different?

If a process is the same in your company as in every competitor, buy it. Payroll, accounting, email, calendars, storage. Nobody wins a market with a better general ledger, and building one costs you the money that should have gone into the thing that does differentiate you. The exceptions are the processes where your way of doing it is the reason customers choose you. That is usually one or two processes, not ten, and most companies can name theirs in a sentence.

Question two: does the product fit, or does your business have to bend?

Every purchased product asks you to work its way. Sometimes that is a gift, because its way is better than yours. Sometimes it is a slow tax: workarounds, a shadow spreadsheet beside the tool, a person whose job is to reconcile the two. The workaround is the signal. If the first thing your team did after rollout was build a spreadsheet next to it, the product does not fit, and no amount of configuration will change that.

Question three: what does the seam cost?

This is the question comparisons leave out. Bought products rarely stand alone; the cost is in the seams between them. Data has to move, definitions have to agree, and somebody has to notice when a sync fails. Three good products with two bad seams is worse than two adequate products with one good seam. When counting the price of buying, count the seams, because that is where the manual work reappears after the migration everyone celebrated.

Question four: who is on the hook in year two?

A purchased product comes with a vendor who fixes it and a roadmap you do not control. A built system comes with no vendor and a roadmap you do control, which is an advantage only if someone is actually operating it. That is the real question behind build versus buy, and it is why my ladder ends in an operator retainer rather than a handover: building without operating just relocates the problem to month six.

The comparison most people do not run

A mid-sized seat-priced product against a one-time build, over three years, for a process that will not go away:

Bought, 20 seats at $40 per month	$9,600 per year, $28,800 over three years
Plus the seam	the person reconciling it, usually a few hours a week
Plus the workaround	the spreadsheet beside the tool, unmeasured
Built, fixed scope	from $6,900 once, then nothing to rent
Plus operating it	from $2,900 a month if you want it run for you
Where buying clearly wins	when the product fits and there is only one seam
Where building clearly wins	when the process is how you are different, or seats keep growing

This is deliberately not stacked in favour of building. For most processes the bought option wins this table outright, which is exactly why the teardown will often tell you to buy something rather than hire me.

Common questions, answered straight

Is there a size below which building never makes sense?

Not a headcount, but a threshold: if the process does not cost you at least the price of the build within about two years, buying or leaving it alone is the better answer. The teardown puts a number on that so it is arithmetic instead of instinct.

What about buying and customising heavily?

The most expensive of the three options more often than people expect. Heavy customisation gives you the constraints of a product and the maintenance burden of a build, and it breaks on upgrades. If you find yourself planning it, the question is usually whether the product fits at all.

Can we start bought and build later?

Yes, and it is often the right sequence. Buy to learn what you actually need, then build the one process that turned out to be yours. What makes this expensive is not the sequence but undocumented workarounds, so keep notes on every one you invent.

What do you recommend most often?

Buying, honestly. In most teardowns at least one of the three time sinks is best solved by an off-the-shelf tool, and the analysis says so by name. I only build the part that is genuinely yours.

Which of your processes is actually yours?

Free systems teardown: your three biggest time sinks, priced, and for each an honest call on buy, build or leave alone. In writing, within 72 hours, no obligation.

Get your free systems teardown

Build or buy,decided honestly