The no-code vendor lock-in escape plan

Vendor lock-in is a slow-moving cost. You don't notice it on day 30, you notice it on day 730 — when a customer asks for a SOC 2 control your platform doesn't support, when an investor diligences your stack, or when the platform doubles its prices in a single email. By then the cost of leaving feels enormous, which is precisely how lock-in works.

Four signs you've stayed too long

1. 1Your hosting bill grows linearly with users, not with usage. If 10× the users costs 10× the platform fee, you're paying for the abstraction, not the work.
2. 2You've hit a customisation wall and your team is now writing 'Bubble JavaScript' or 'PowerApps Power Fx' that nobody else in the world hires for.
3. 3An audit (SOC 2, HIPAA, GDPR, PCI) is on the calendar and your platform's compliance docs end at 'we are SOC 2 — but you'll need to demonstrate your controls separately.'
4. 4Your engineering team has stopped applying for jobs because the only thing on their resume is your no-code platform.

The 90-day escape plan

Most teams escape in 90 days, end-to-end, without freezing the product. The plan has three 30-day phases:

Days 1–30: contain

• Inventory the no-code app: pages, workflows, data types, integrations, custom code blocks.
• Lock down further proprietary features — no new plugins, no new platform-specific patterns.
• Stand up a parallel Postgres database with idempotent daily syncs from the no-code source.

Days 31–60: rebuild

• Re-platform the highest-pain module first (usually billing, search, or the admin panel).
• Wire it up to the synced Postgres so the no-code app and the new code see the same data.
• Run both versions in parallel — power users on the new module, everyone else on the old one.

Days 61–90: cut over

• Migrate remaining modules, one per week, behind feature flags.
• Flip DNS, freeze writes to the no-code app, run a final delta sync.
• Decommission no-code 30 days later, after one full billing cycle of stability on the new stack.

What you keep

The product, the brand, the URL, the customers, the team's product instincts, and the lessons baked into the no-code app. You only lose the platform fee and the ceiling.