We sell custom software, so you might expect our answer to build vs buy to be "build". It is not. Roughly half the discovery engagements we run end with us recommending an off-the-shelf product, sometimes with a thin custom layer around it. A wrong build wastes a year. A wrong buy warps your operations around someone else’s roadmap. Both are expensive, so the decision deserves better than instinct.
The one question that settles most cases
Is this capability part of how you win, or part of how you run? Payroll, accounting, email, HR leave tracking: every competitor has them, no customer chooses you for them. Buy them, always. The process that makes customers pick you over the alternative, the thing you do differently on purpose: that is where owning the software can compound into an advantage.
Most companies have two or three genuinely differentiating processes and dozens of commodity ones. Teams that get this wrong usually err in one direction: they believe their commodity process is special. "Our invoicing is unique" is almost always false. "Our credit-decisioning model is unique" might be true, and it is worth a hard look.
Score it before you argue about it
| Factor | Points to build | Points to buy |
|---|---|---|
| Differentiation | Core to why customers choose you | Same as every competitor |
| Fit of existing tools | Best option covers under 70% of the need | A product covers 90%+ out of the box |
| Rate of change | Process changes monthly, you need control | Process is stable and standard |
| Integration depth | Must sit inside your core data and workflows | Happy at the edge with an export |
| Total 5-year cost | Licences exceed a build over 5 years | Licences are a fraction of a build |
| Capacity to own it | You have or want product ownership muscle | Nobody will own it after launch |
Score each row honestly with the people who live in the process, not just the ones who will present the decision. Four or more in one column is a clear answer. A split usually means the scope mixes a commodity workflow with a differentiating one, and they want different answers.
The hybrid answer that wins most often
The strongest pattern we see in mid-size companies is buy the commodity, build the edge. Buy accounting, but build the quoting engine that prices jobs the way only you can. Buy a support desk, but build the fulfilment tracker your customers actually log into. Lumina Retail runs on exactly this shape: off-the-shelf finance and HR, a custom stock-allocation engine at the centre, and thin integrations stitching them together. The custom surface is small, which keeps the owning cost small, and every custom line of code sits where it earns margin.
- 1List the workflows in scope and label each one: differentiating or commodity.
- 2For every commodity workflow, shortlist two products and check real fit against your ten most common cases, not the demo script.
- 3For the differentiating core, scope the smallest build that owns the workflow end to end.
- 4Design the seams: where data flows between bought and built, and who owns each integration.
Traps on both sides
- Building because the demo disappointed. A tool that covers 85% of the need with configuration beats a build that covers 100% in eighteen months.
- Buying into deep customisation. If implementation involves months of consultant-written custom code inside someone else’s platform, you are building anyway, with worse tools and a landlord.
- Ignoring exit costs. Ask how your data comes out before you put it in. On the build side, ask who maintains the system in year three before you write line one.
- Deciding once, forever. Revisit the call every couple of years. Products mature, your edge moves, and last year’s right answer can quietly expire.
Deciding this week
You do not need a quarter-long evaluation. Label the workflows, score the table, price both sides over five years, and pressure-test the result with someone who has run both kinds of project. We run that exercise with clients as part of discovery, and we put the recommendation in writing even when the recommendation is "do not hire us to build this."