How a Jakarta SaaS Founder Survived a Platform Migration with a Fractional CTO

The founder of a 22-person B2B SaaS based in Jakarta called us in a state we recognised. He’d just received a proposal for a complete platform rebuild from his existing development agency. Eight hundred and twenty million Rupiah, nine months, with the agency’s promise that “this is the only way to fix the architectural issues.”

He had two questions: was this real, and was there an alternative? The reason he was calling us, not committing, is that something didn’t smell right. He couldn’t articulate what.

Twelve weeks later, the rebuild didn’t happen. The platform was stable, the team was shipping, and total spend was Rp 95 juta — not Rp 820 juta. Here’s how that decision got made.

The starting picture

A two-week diagnostic produced this view:

• The platform was 4 years old, originally built by the same agency. Some real architectural debt, but not catastrophic.
• Customer-facing reliability was OK. 99.4% uptime over the previous 6 months. Not great, not terrible.
• Performance was getting slow in specific areas — particularly the reporting module — but not the whole platform.
• The agency’s pitch was framed as “incremental fixes are throwing money away; only a rebuild fixes this properly.”

Underneath the agency’s pitch we found three things that changed the picture:

1. The “architectural issues” were largely confined to one module. The reporting subsystem had a fundamentally bad design that did need rebuilding. The other 80% of the platform was fine.
2. The agency’s proposed rebuild was the same architecture they used for their other clients. Not custom-fit to this SaaS — a templated approach that would have created a different set of problems within 18 months.
3. The agency’s monthly maintenance retainer had grown from Rp 12 juta to Rp 28 juta over two years. No clear scope expansion. Nobody had questioned it.

The founder had been operating without senior tech judgment for the entire 4 years. The agency had become both his vendor and his technical advisor — a setup with predictable conflicts of interest.

The intervention

A 12-week sprint engagement with a clear scope:

Weeks 1–3: Independent assessment

A senior engineer (not us — a third-party we trust for this kind of work) audited the actual codebase. Their report: “Reporting module needs rebuilding. Other modules need targeted refactoring on three specific bottlenecks. No need for full platform rebuild.”

This single piece of evidence was the inflection point. The founder could now push back on the agency with a concrete alternative diagnosis.

Weeks 4–6: Vendor renegotiation

We helped the founder restructure the agency relationship. Three changes:

• Maintenance retainer reset to Rp 18 juta/month with explicit scope (security patches, dependency updates, 25 hours of feature work). The previous Rp 28 juta had been creep with no scope.
• Reporting rebuild scoped at Rp 95 juta over 10 weeks — a focused project, not a full platform rebuild. Fixed price, milestone-based payments.
• Code ownership and access transferred to the founder’s accounts. He’d never had it before; the agency had been controlling production access.

The agency pushed back hard on these changes for three weeks. The fractional CTO held the line because the alternative (the Rp 820 juta rebuild) was clearly worse for the SaaS.

Weeks 7–10: Reporting rebuild

The targeted reporting rebuild proceeded. Came in at Rp 89 juta, slightly under quote. Performance issues in the reporting module dropped by 80%. The other parts of the platform remained untouched and continued working.

Weeks 11–12: Future state and handover

Built a written technology roadmap for the next 18 months: when to address the next architectural concerns, what hires to consider, vendor contract review schedule. The fractional engagement ended; the founder retained the option for an ongoing advisory retainer at Rp 18 juta/month for 4 hours of availability per week.

What it cost vs what it saved

Total fractional CTO engagement: Rp 95 juta over 12 weeks.

Direct savings: the avoided Rp 820 juta rebuild. Indirect savings:

• Maintenance retainer reduction from Rp 28 juta to Rp 18 juta = Rp 120 juta/year ongoing
• Reporting module fix at Rp 89 juta vs the agency’s allocated Rp 200 juta within the rebuild
• Code ownership transferred — eliminates future lock-in cost of 6+ figures

Total first-year savings, conservatively: Rp 600 juta+. Ongoing annual benefit: Rp 120 juta.

What we got wrong along the way

Two things, worth being honest about.

The first: we underestimated the agency’s resistance. They had been the founder’s technical advisor for four years. Restructuring that relationship took longer than we’d budgeted — three weeks of difficult conversations rather than one. We didn’t lose the engagement, but it strained the founder’s energy more than it should have.

The second: the third-party assessment in weeks 1–3 should have been the agency’s competitor, not a neutral senior engineer. We chose neutral to avoid bias. In hindsight, having an alternative vendor’s perspective would have given the founder more leverage in the renegotiation.

What we’d recommend to similar founders

Three patterns from this case that generalise:

1. If your existing vendor is also your technical advisor, you don’t have a technical advisor. The conflict of interest is too structural. A fractional CTO solves this without requiring you to fire the vendor.
2. Big rebuild proposals deserve independent review. The cost of a 2-week independent assessment is 1–2% of the proposed rebuild. Always worth it.
3. Vendor retainer creep is real and quiet. Compare your current monthly retainer to what it was two years ago. If it’s grown without clear scope expansion, that’s worth questioning.

If you’re staring at a major technology decision and want an outside perspective before committing, an hour of conversation usually clarifies whether the case is similar to this one. We do those at no cost.