"One of the biggest mistakes in software is continuing failed products. Define clear limits."
FRAMEWORK STATUS: ENFORCED • REVISION 2.4 • SEPTEMBER 2026
Conventional incubators and agencies frequently succumb to the sunk cost fallacy—investing continued engineering hours into stagnant products simply because capital was already spent. At XPELAB, we consider ending an underperforming experiment a measure of operational success, not failure. Killing an unvalidated idea preserves engineering bandwidth and redirects capital toward initiatives with high market adoption.
Every software initiative incubated in our lab must successfully clear five sequential empirical gates before qualifying for subsequent funding stages:
Do customers actually experience this pain frequently? If NO → Stop experiment immediately.
Will target businesses test an interactive prototype or pilot workflow? If NO → Pivot or stop.
Will customers commit hard capital to solve the problem? If users like the product but won't pay, the business model is unviable.
Do pilot users continue active utilization? We monitor monthly active usage (MAU), recurring log-ins, and churn rates.
Can customer acquisition and infrastructure scale while recurring revenue grows faster than operational costs? If YES → Allocate scale capital.
Following initial prototype or pilot deployment, every incubated experiment operates on a strict 60-day evaluation countdown with predefined minimum viability quotas:
The hypothesis is validated with paid demand. Allocate Phase 2 expansion budget, productize features, and integrate into regular release cycles.
High user engagement with friction in payment checkout. Granted one 30-day pivot cycle to re-align pricing or core workflows.
Shut down active servers, notify pilot participants, extract generic architectural code into reusable libraries, and reassign developers.
Capital allocations are strictly tiered. Engineering leads may not bypass a budget tier without meeting the preceding gate:
| Level | Stage | Budget Ceiling | Exit Deliverables |
|---|---|---|---|
| Level 1 | Idea Validation | ₹2,000 – ₹5,000 | Customer interview notes, landing page test. |
| Level 2 | Interactive Prototype | ₹5,000 – ₹20,000 | Clickable wireframes, core UI workflows. |
| Level 3 | Core MVP | ₹20,000 – ₹75,000 | Auth, database, payments, working admin panel. |
| Level 4 | Pilot Testing | ₹50,000 – ₹2,00,000 | 10+ live users, feedback logs, performance fixes. |
| Level 5 | Product Scale | Funded by Revenue | CAC < LTV, churn < 5%, positive unit economics. |
When an experiment is terminated, its development time is never wasted. The engineering squad executes our Code Harvest Protocol:
Our review board convenes on a continuous 4-week cadence to evaluate all active lab experiments: