Build vs buy your factory software

If you've spent a few weekends comparing factory-software vendors, you've probably had the same circular thought we hear every week: "the off-the-shelf option doesn't quite fit, but a custom build sounds expensive and risky." Both halves of that thought are usually right at the same time.

This is the framework we walk through with clients on discovery calls when they're stuck on this decision.

The four real options (not just two)

The mental model most people start with is binary: buy SaaS, or build custom. The reality is four:

1. Pure off-the-shelf SaaS. Pay per user per month. Configure to taste. The vendor owns everything underneath.
2. Off-the-shelf with heavy customisation. A platform vendor with a customisation services arm bills you to extend their product to fit. Usually the most expensive of all four — you pay enterprise services rates on a vendor-controlled stack.
3. Hybrid: bought platform + custom integration layer. Buy a strong off-the-shelf component (an accounting package, a warehouse management system) and have someone build the custom integrations and process software around it.
4. Pure custom build. Software designed for your factory, owned by you, hosted where you choose. We do this when it's the right answer.

The hybrid (option 3) is what most successful SMB factories actually end up with. Almost nobody talks about it because it doesn't fit a single vendor's pitch.

When pure off-the-shelf SaaS is right

Concrete signals:

• The process is genuinely standard. Accounting is accounting. Email is email. CRM is CRM (mostly). If your process is the same as 10,000 other businesses, paying a SaaS vendor to maintain it for you is the right answer.
• Per-seat pricing makes sense for your team size. SaaS at RM 80/user/month for 10 users is RM 9,600/year. Same SaaS for 200 users is RM 192,000/year. The math flips at scale.
• You don't care about exit risk. If you can switch providers in 6 months without much pain, vendor lock-in is tolerable.
• Out-of-the-box reports cover what you need. If you're already exporting to Excel for "real" reporting, the SaaS isn't actually doing the job.

For SMB Malaysian factories, pure SaaS is usually right for: accounting, payroll, HR, basic CRM, email/collaboration. It's usually wrong for: production floor, OEE, custom workflows, anything industry-specific.

When pure custom is right

Concrete signals:

• Your process is your competitive advantage. Owners often forget this — but the way your factory routes orders, splits batches, handles claims, deals with returns is often the unique thing customers value. Buying generic SaaS that flattens it into industry-standard moves is sometimes a quiet way to lose your edge.
• Off-the-shelf doesn't exist for what you need. OEE platforms exist. ERPs exist. But "OEE platform that knows your specific equipment, talks to your specific PLCs, alerts your specific WhatsApp groups, integrates with your specific accounting" — that doesn't exist as a product. Someone has to build it.
• You'll use it for 5+ years. Build cost is one-off. Run cost is ongoing. Over a 5-year horizon, custom typically beats heavy-customised SaaS on total cost.
• You want to own it. Code, data, hosting, schema. Custom means yours. SaaS means rented.

For SMB Malaysian factories, custom is usually right for: production-floor software (OEE, dashboards, alerts), customer-facing portals, mobile apps for floor staff, AI employees, anything that touches your factory's specific equipment and process.

When hybrid is right (most of the time)

The hybrid pattern: buy commodity components, build the custom layer that fits your factory.

A typical SMB hybrid stack:

• Accounting: off-the-shelf (Xero, AutoCount, SQL Account, etc.)
• Email / collab: off-the-shelf (Google Workspace, Microsoft 365)
• OEE / Industry 4.0: custom (no off-the-shelf product fits)
• Production scheduling: custom (your scheduling logic is yours)
• Customer-facing portal: custom (your B2B flow is unique)
• MyInvois integration: custom layer in front of the off-the-shelf accounting
• Mobile apps: custom (your floor staff workflows are yours)
• Reporting: custom dashboards pulling from all of the above

This is the answer for ~80% of the SMB factories we work with. The custom layer is where your competitive advantage lives. The bought layer is where standardisation makes sense.

The mistake most vendors push you to make: try to fit the custom-shaped problems into off-the-shelf-shaped products. The other mistake (less common): try to build the bought-shaped commodities yourself.

Cost comparisons that hold up

Honest 5-year total cost framing for an SMB factory needing OEE + production scheduling + custom mobile app + MyInvois + accounting integration:

Pure off-the-shelf bundle: RM 30,000–RM 80,000/year in subscriptions + RM 50,000–RM 200,000 in mandatory customisations + ongoing customisation fees. 5-year total: RM 250,000–RM 1,200,000. And it still doesn't quite fit.

Pure custom build: RM 200,000–RM 500,000 build + RM 12,000–RM 36,000/year run + occasional extension projects. 5-year total: RM 260,000–RM 700,000. Fits exactly.

Hybrid (bought commodities + custom factory layer): RM 5,000–RM 15,000/year SaaS + RM 100,000–RM 300,000 custom build + RM 8,000–RM 24,000/year run. 5-year total: RM 165,000–RM 450,000. Best value.

Numbers vary widely. The point isn't the exact figures — it's the shape of the comparison. Hybrid usually wins, sometimes by a lot.

Where each option fails

Honest failure modes:

Pure SaaS:

• Slow death by feature mismatch. The vendor's roadmap moves toward bigger customers. Your edge cases get deprioritised. After 3 years you're paying more for less.
• Lock-in surfacing at renewal. Switching costs are revealed when you try to leave.

Heavy-customised SaaS:

• The customisation grows into a parallel custom system on top of the vendor's platform — but at vendor services rates and on the vendor's schedule.
• When the vendor changes their platform, your customisation breaks.

Pure custom:

• Cost overruns when scope isn't clear at start.
• Lock-in to the developer who built it (mitigated by clean code, documentation, your-cloud hosting).
• Slow to build the commodity layers (no good reason to rebuild accounting).

Hybrid:

• Integration complexity at the seams. Each system you bought has its own data model and you have to translate.
• Vendor-vs-developer finger-pointing when something breaks at the boundary. Mitigate by having one team own the integration layer end-to-end.

How to make the decision for your factory

A short framework for an SMB owner:

1. List every system you'd need. Production floor, accounting, payroll, customer portal, internal admin, mobile apps, reporting. About 8–12 systems for a typical SMB.
2. Mark each as Standard or Specific. Standard = same as 10,000 other businesses. Specific = your factory has its own way.
3. Standard ones get bought. Don't rebuild what already exists at SaaS prices.
4. Specific ones get custom. That's where competitive advantage lives.
5. Plan the integration layer separately. This is where Standard meets Specific. Get it owned by one team.
6. Phase the build. First year: get OEE + the most painful Specific layer working. Year two: extend.

Where we sit

We do option 3 (hybrid) and option 4 (custom) for Malaysian SMB factories. We don't do option 1 — there are vendors who specialise in selling SaaS and we're not them. We don't do option 2 either — heavy-customised SaaS is typically the worst-value option and we'll usually advise against it.

If you're stuck on the build-vs-buy decision for your factory and want a 30-minute walkthrough of what the right shape would be, drop us a line — free discovery call, no obligation, honest answer.