When 'build' is the right answer
Building makes sense when your procurement workflow is genuinely unique to your industry, you have a stable in-house engineering team with bandwidth, and the workflow won't need to change for 3+ years.
Examples where building can succeed: defense contractors with classified-supplier workflows, regulated medical device companies with FDA traceability needs, and a few specialty chemicals manufacturers.
When 'build' is a trap
Building is almost always a trap when the goal is 'we want it to fit our exact process.' That goal sounds reasonable and is genuinely the most expensive sentence in enterprise software.
Building also collapses when the engineering team that built it leaves.
The real 3-year TCO of building
For a 250-employee company, an honest internal build runs:
- Year 1: $480k–$720k (2 engineers + PM + design)
- Year 2: $320k–$480k (maintenance + new features)
- Year 3: $320k–$480k (maintenance + integrations)
- 3-year total: $1.12M–$1.68M, plus opportunity cost of engineering capacity
Hypothetical example.
How to decide in one meeting
Ask three questions: (1) Is our procurement workflow truly unique, or do we just think it is? (2) Will we still have the engineering team in 3 years to maintain this? (3) What is the highest-ROI thing those engineers could be building instead?
If the answer to any of those is uncertain, buy.
