Every growing company reaches the same argument. Someone wants to buy a platform. Someone else says it is cheaper to build. Both are usually wrong, because the argument is about preference rather than arithmetic. Here is a way to settle it with numbers in about an hour.
What buying really costs
The subscription is the visible part. The rest is setup, configuration, the annual increase, the seats you keep paying for after people leave, integration work and the cost of changing your process to match the tool.
| Cost line | Typical shape |
|---|---|
| Subscription | Per user per month, often rising 5 to 10 percent a year at renewal |
| Implementation | One time, often 30 to 100 percent of year one subscription |
| Integration | One time plus maintenance whenever the vendor changes their API |
| Process change | Training, workarounds, and the hours your team spends fitting the tool |
| Switching | Real, and usually ignored. Data export, contract terms and re-training |
What building really costs
Building is not one payment. It is discovery, design, build, launch and then the part people forget: maintenance. Software needs updates, security patches, dependency upgrades and small changes as the business shifts. Budget 15 to 25 percent of the build cost each year for that, whether you do it in house or with a partner.
When buying wins
- The process is standard, such as payroll, accounting or email.
- The vendor carries real regulatory weight, such as tax filing changes.
- You need it live in weeks, not months.
- Your team has no capacity to own software long term.
When building wins
- The process is how you compete, not how everyone works.
- You are paying for a large platform to use one small feature.
- Your data model is unusual enough that every tool forces a workaround.
- You already have engineers who can own it.
A scorecard you can run
| Question | Points for build |
|---|---|
| Is this process a source of advantage for us? | 0 to 3 |
| Do at least three vendors fail on a requirement we cannot drop? | 0 to 2 |
| Is the workflow stable enough to be worth automating? | 0 to 2 |
| Will we still need this in three years? | 0 to 2 |
| Do we have, or can we hire, someone to own it? | 0 to 1 |
Score 7 or more and building deserves a serious look. Score 3 or less and buying is almost certainly the right call. The middle is where a paid blueprint earns its keep: two weeks of scope and a fixed build price lets you compare a real number against the subscription quote instead of a guess.
The hybrid most companies land on
Buy the commodity, build the edge, and connect the two properly. Payroll stays bought. The pricing engine that wins you deals gets built. The integration between them is where the value leaks if nobody owns it. Our software team and business systems team are set up so you can do both with one group of people.