We are a software house, so recommending that you buy an off-the-shelf product costs us revenue. We recommend it anyway when it is the right answer, because building something that should have been bought costs the client far more — and eventually costs us the relationship.
Here is the framework.
Question 1: is this process a differentiator or a commodity?
Payroll is a commodity. So is email, accounting, and in most cases CRM. Nobody wins customers by having a better payroll process.
Your pricing logic, your fulfilment model, the specific way you serve your particular customers — those may well be differentiators.
Buy commodities. Build differentiators. This single question resolves most cases.
Question 2: how much would you have to change to fit the product?
Try the product with your actual process. If the gap is "we would adjust two steps", buy it — your process was probably not optimal anyway.
If the gap is "we would have to stop doing the thing our customers choose us for", that is the signal to build.
Be honest here, because there is a bias in both directions: teams overstate how special their process is, and vendors understate how much you would have to change.
Question 3: what does the total cost look like over five years?
Buying: licence × users × 60 months, plus implementation, plus integration, plus the annual increases in the contract you have not read carefully.
Building: development cost, plus roughly 15–20% of that annually for maintenance and change.
For a 15-person team on a $60/user/month product, five years is around $54,000 in licence alone. A custom build at $40,000 with $7,000 a year of maintenance comes to $75,000. Buying wins.
At 80 users, the same product is $288,000 over five years and the build is unchanged. Building wins decisively.
Headcount growth is the variable that flips this, and it is the one most often left out of the comparison.
Question 4: is there a hybrid?
Frequently the best answer is neither. Buy the commodity platform, build the differentiating layer on top, integrate them properly.
A typical case: a manufacturer is about to build a full ERP because its production planning is genuinely unusual. The better answer is usually Odoo for finance, inventory and purchasing, with a custom production planning module built against its API. The differentiator gets built, the commodity gets bought, and the bill is a fraction of a full build.
Question 5: what happens if the vendor disappears or triples the price?
For a commodity, the answer is "we migrate, painfully but survivably". For the system that runs your core operation, "we are out of business" is a genuine possibility, and it is an argument for building or for insisting on data portability and an escrow arrangement.
The honest summary
Most businesses should buy more than they build, and build a narrower thing than they think. The instinct to build everything usually comes from a bad experience with a product that did not fit — which is often a selection failure rather than a category failure.
If you want an unbiased read on your specific case, we will give you one, including when the answer is that you do not need us.