Contents
- A 2022 Stripe Developer Report found that the average engineering team spends 33% of its time on infrastructure maintenance and integration work -- tasks that have nothing to do with the product they are building.
- Stealth Labz deployed 10 production systems across 7 verticals using a shared architecture called a scaffold.
- Every product built on the shared architecture contributes back to it.
Over 80% of software infrastructure transfers directly when you launch a new product vertical -- if the original system was built with reuse in mind. The transferable portion includes authentication, database patterns, admin interfaces, API routing, deployment pipelines, analytics frameworks, and notification templates. The remaining 20% is the vertical-specific business logic, market-specific customization, and any geography-dependent compliance rules.
Why most companies don't get this number
A 2022 Stripe Developer Report found that the average engineering team spends 33% of its time on infrastructure maintenance and integration work -- tasks that have nothing to do with the product they are building. That overhead recurs with every new product because the infrastructure was built for one purpose and doesn't travel well. Monolithic architectures, vendor lock-in, and tightly coupled business logic all prevent reuse.
The result is what developers call the "cold-start tax": the weeks and thousands of dollars spent setting up frameworks, environments, authentication, databases, and deployment pipelines before a single line of product-specific code gets written. Traditional cold-start costs run 24 to 36 days and $5,000 to $15,000 per new product.
What 80%+ transfer looks like in practice
Stealth Labz deployed 10 production systems across 7 verticals using a shared architecture called a scaffold. Here is what transferred versus what required new work across four insurance sub-verticals (life, auto, annuities, financial services):
Transferred automatically (80%+):
- Authentication and role management
- Database schema and migrations
- Admin interface components
- API structure and routing
- Deployment pipeline
- Error handling and logging
- Analytics framework
- Email and notification templates
Required vertical-specific development (approximately 20%):
- Vertical-specific business logic (e.g., life insurance quoting vs. auto insurance quoting)
- Market-specific customization (carrier integrations, product-specific user experience)
- Geography-specific compliance rules
The fourth product in the insurance cluster (financial services) took only 11 active development days -- compared to 24 days for the first -- because the scaffold deepened with each deployment. The infrastructure got richer, not thinner, as more products drew from it.
The compounding effect
Every product built on the shared architecture contributes back to it. Authentication patterns proven in one vertical deploy instantly to the next. Payment processing logic built for an e-commerce product transferred into the insurance quoting cluster. Lead capture flows designed for insurance redeployed to legal and reporting verticals. The scaffold carries its quality history forward -- 3.7% defect rates in the insurance cluster came from deploying battle-tested patterns, not from extraordinary effort on each individual build.
Related: How much cheaper is shared infrastructure vs. separate builds for each product?
References
- Stripe (2022). "Developer Coefficient." Survey data on engineering time spent on infrastructure maintenance and integration work.
- Keating, M.G. (2026). "Case Study: The Scaffold." Stealth Labz. Read case study
- Keating, M.G. (2026). "The Compounding Execution Method: Complete Technical Documentation." Stealth Labz. Browse papers