Contents
The Claim
CEM claims that execution compounds — that each project doesn't just produce a product, it produces reusable infrastructure that makes subsequent projects faster, cheaper, and higher quality.
This case study puts numbers on that claim.
The Evidence: Three Curves That All Move Together
Curve 1: Speed Increased
Days to Ship a Functional Product
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Project 1 ████████████████████████████████████████████ 24 days
Project 2 ████████████████████████████████████████████ 25 days
Project 3 ████████████████████████████████████████ 23 days
Project 4 ██████████████████████████████████████████████████ 28 days (7 integrations)
Project 5 ████████████████████ 11 days
Project 6 ████████████████████████████████ 20 days (new geography)
Project 7 ████████████████ 16 days
Project 8 █████████ 9 days
Project 9 █████ 5 days
──────────────────────────────────────────
24 days → 5 days. Same operator.Curve 2: Cost Decreased
External Support Cost Per Project
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Project 1 ████████████████████████████████ $7,995
Project 2 ████████████████████ $4,080
Project 3 ████████████████████ $4,005
Project 4 ████████████████ $1,680
Project 5 (included in flagship)
Project 6 ███ $330
Project 7 ██ $330
Project 8 █ $90
Project 9 ▏ $0
──────────────────────────────────────────
$7,995 → $0. Same operator.Curve 3: Quality Improved (or Held)
Product Defect Rate by Project
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Industry norm ████████████████████████████████████████ 20-50%
Scaffold cluster ████ 3.7-3.9%
Financial services ███████████ 11.3%
Reporting platform ████████████████ 16.1%
Seasonal e-commerce ████████████████████ 16.8%
Portfolio average ████████████████████████ 23.7%
Flagship (complex) ████████████████████████████████ 31.3%
Even the worst project is within industry norms.
The best projects are 5-10x better.Three curves. All moving in the right direction. Simultaneously. That's not random variation — it's a compounding system producing measurable returns.
How Foundation Actually Works
Every project produces two outputs:
What Each Project Creates
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
┌──────────────────────────────────────┐
│ THE PRODUCT │ ← What the customer sees
│ Features, interfaces, logic │
└──────────────────────────────────────┘
┌──────────────────────────────────────┐
│ THE FOUNDATION │ ← What the next project inherits
│ │
│ Authentication patterns │
│ Database schemas │
│ Admin interface components │
│ API structures │
│ Deployment pipelines │
│ Error handling patterns │
│ Integration templates │
│ Quality patterns │
└──────────────────────────────────────┘The product is the deliverable. The foundation is the investment. Every project makes the foundation deeper, and the deeper the foundation, the less work the next project requires.
A Concrete Example
How Authentication Traveled Across Projects
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Project 1: Built authentication from scratch.
Multi-tenant roles: Admin, Partner, Affiliate, Business.
Cost: days of work + external support.
Project 3: Inherited authentication from Project 1.
Added insurance-specific role permissions.
Cost: hours, not days.
Project 5: Inherited authentication + insurance roles.
Customized for different geography.
Cost: minimal — pattern was proven.
Project 9: Inherited everything.
Authentication deployed in minutes, not days.
Cost: effectively zero.This happened with every reusable component — database patterns, admin UIs, API structures, deployment configs. Each one followed the same curve: expensive to build the first time, nearly free to deploy the tenth time.
The Compounding Math
Template Reuse: 95%+
By the late portfolio, 95%+ of infrastructure components came from foundation. New projects only required the logic that made them different — typically 5-20% of total functionality.
Work Required Per Project Over Time
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Project 1 ░░░░░░░░░░░░░░░░░░░░████████████████████████ New work: ~90%
Project 3 ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░████████████ New work: ~30%
Project 5 ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░██████ New work: ~15%
Project 9 ░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░███ New work: ~5%
░ = From foundation █ = New work requiredWhat This Means in Business Terms
| If you're building... | Without Foundation | With Mature Foundation |
|---|---|---|
| 1 product | Full build cost | Full build cost (same) |
| 3 products | 3x build cost | ~1.5x build cost |
| 5 products | 5x build cost | ~2x build cost |
| 10 products | 10x build cost | ~3x build cost |
Foundation makes the economics of multi-product businesses radically different. The marginal cost of each new product decreases toward zero as the foundation deepens.
The Cross-Pollination Effect
Foundation doesn't just transfer within a product family — it transfers across verticals and geographies:
How Foundation Flows Across the Portfolio
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Insurance (Life) ──┐
Insurance (Auto) ──┤── Insurance patterns ──┐
Insurance (Annuities)┤ │
Insurance (Finance)──┘ │
├── Combined foundation
Operations platform ── Admin/CRM patterns ───┤
│
E-commerce ── Checkout/payment patterns ─────┤
│
South Africa ── Multi-geo patterns ──────────┘
│
▼
Every new project draws from ALL of thisWhen the US insurance quoting product shipped in 16 days at $330, it wasn't just drawing from the South African version. It was drawing from the insurance cluster's vertical patterns AND the US-based products' geography patterns AND the flagship platform's admin patterns. Foundation compounds across every axis simultaneously.
Why It Matters
Compounding is the most powerful force in business — but most people only think about it in financial terms. CEM applies compounding to execution. Every project is simultaneously a product delivery and an investment in future speed, quality, and cost reduction.
This is why "build vs. buy" calculations are wrong. Traditional analysis compares the one-time cost of building to the ongoing cost of buying. It misses the foundation effect — the fact that building product #1 makes products #2–10 dramatically cheaper to build. The real comparison isn't "build one product vs. buy one platform." It's "build a foundation that produces 10 products vs. buy 10 platforms."
The operator who builds early compounds longest. Foundation depth is a function of projects shipped. The earlier you start building, the deeper your foundation grows, and the faster and cheaper everything that follows becomes.
Key Numbers
| Metric | Value |
|---|---|
| Template reuse (mature) | 95%+ |
| Build time compression | 24 days → 5 days |
| Cost compression | $7,995 → $0 |
| Portfolio size feeding foundation | 10 projects |
| Cross-vertical transfer | Confirmed across insurance, e-commerce, operations, legal |
| Cross-geography transfer | Confirmed across US and South Africa |
References
- McConnell, S. (2004). Code Complete, 2nd ed. Microsoft Press. Industry defect rates of 20–50% for typical software projects.
- Keating, M.G. (2026). "Foundation: The Compounding Knowledge Base." Stealth Labz CEM Papers. Read paper
- Keating, M.G. (2026). "Scaffold: Deployable Architecture as Execution Accelerator." Stealth Labz CEM Papers. Read paper