Build vs Buy vs Configure: A Decision Framework for Growing Companies

Every software need has three answers, not two. Most decisions get framed as “build vs buy”, which is the wrong frame because it skips the third and often-best option: configure existing tools to do most of what you need.

Here’s the framework that consistently produces the right answer.

The three options, plainly

Buy: Subscribe to off-the-shelf SaaS. Use it as designed. Minimal customisation. Pay monthly, forever.

Configure: Take an existing platform — could be a SaaS like HubSpot, an open-source tool like Metabase, or a no-code platform like Airtable — and configure it to fit your specific workflow. You might write a small amount of custom code or scripts, but the bulk of the work is configuration, not engineering.

Build: Write custom software from scratch (or close to it). Full control, full ownership, full development cost.

These aren’t always mutually exclusive — many systems combine all three. The framework helps you decide which dominates a given need.

The decision questions

Five questions, in priority order:

1. Is there a SaaS that does this well?

If the SaaS market has 5+ mature options that all do roughly what you need, buying is almost always right. Don’t reinvent. Generic workflows have generic solutions, and the market is brutal enough that the surviving SaaS products are good.

If the answer is yes and the SaaS is affordable for your scale, you can stop here.

2. If yes, are you paying for things you don’t use?

Sometimes the SaaS is fine but the tier you need locks you into paying for unused features. Common pattern: you need feature X, which only comes in the “Enterprise” tier at 5x the price.

When this happens, you have two choices:

• Eat the cost (often correct if your need is temporary or marginal)
• Configure a cheaper alternative (often correct if your need is permanent and the math works)

Configure usually beats build here. Most workflows that look custom can be assembled from a SaaS plus configurations.

3. Can you configure an existing platform to do 80% of it?

If yes, configure. Common combinations that work:

• HubSpot for CRM with workflows + custom fields + light Zapier glue
• Metabase or Looker Studio for dashboards on top of your existing data
• Airtable for operational tools that need to be flexible
• Notion or Coda for internal docs and lightweight workflows
• Make or n8n for workflow automation between SaaS tools

Configuration projects typically cost Rp 10–60 juta and ship in 2–6 weeks. Build projects in the same complexity range cost Rp 80–250 juta and ship in 8–16 weeks. The configured version is also easier to change later.

4. If configure won’t reach 80%, is the gap worth building?

Sometimes the configurable platforms get you to 60% and the missing 40% is core to your business. That’s when build wins.

Test: if you wrote down the missing 40% as “things our system needs to do that no SaaS does”, how many of them are actually critical? If it’s 1–3, build a small custom system that handles just those. If it’s 10+, you’re heading into a real custom build.

5. If you build, is it cheaper than the SaaS over 3 years?

Calculate honestly:

• SaaS: monthly cost × 36 months + integration time
• Configure: build cost + configuration time + ongoing platform cost × 36 months
• Build: build cost + maintenance × 3 years + hosting × 3 years

For most SME workflows, configure wins on both cost and time-to-value. Build wins when the workflow is genuinely unique and high-value enough to justify the lifetime cost.

The trap most SMEs fall into

Two common mistakes:

Building when configure would have worked

Companies decide they need custom software because the off-the-shelf options “don’t quite fit”. They never seriously evaluate whether configuration could close the gap.

In our experience, about 40% of “we need to build custom” projects could have been solved with configuration of HubSpot, Metabase, Airtable, or similar — at a fraction of the cost and time. Always run the configure exercise before committing to build.

Buying then breaking the SaaS to make it fit

The opposite mistake: buy a SaaS, then customise it heavily through fragile workarounds. Within a year you have a Frankenstein version of HubSpot held together with Zapier and prayers, that costs more than custom would have and breaks every time the SaaS updates.

If you’re customising the SaaS more than 30% beyond its design, consider configuring a different platform or building.

A working pattern for SME stacks

The shape that consistently delivers:

1. Buy generic stuff. Email, accounting, payroll, file storage. Don’t think about it.
2. Configure operational tools. CRM, dashboards, internal workflows. Pick a platform that fits the shape, configure it.
3. Build the 5–15% that’s actually unique to you. Custom only the parts where you have a genuine competitive reason.

Most SMEs that get this right have 80% buy, 15% configure, 5% build. Most SMEs that get it wrong have 40% buy, 5% configure, 55% build — and they’re paying for the difference in time and money for years.

If you’re approaching a software need and want to stress-test which of the three options actually fits, an hour of conversation usually clarifies it. We do those at no cost.