No Free Lunch: The Debt, The Excuses, and The Reality (Ch 3)

Chapter 3 confronts arguments against data modeling (“too much work,” “AI can do it,” “we’re moving too fast”) and introduces the three-debt framework:

1. Technical debt — short-term code/schema decisions that compound
2. Data debt — poor quality, governance, documentation, and models accumulating defects
3. Organizational debt — eroded trust, siloed teams, lost credibility from bad data

These feed each other in a “Compounding Debt Loop”: messy code creates fragile models, fragile models produce conflicting metrics, conflicting metrics destroy trust, eroded trust forces more quick hacks. Uses the e-commerce JSON column case study to trace all three debts from one shortcut.

Reis frames data modeling as a spectrum between fast/reckless and slow/rigorous, with the ideal being intentional modeling calibrated to constraints. Exceptions: MVPs, exploratory analysis, and legacy/third-party systems where you have no control.

RDCO relevance

The three-debt framework is a powerful consulting tool. We can help clients quantify their debt — especially organizational debt (the “punch pass” metaphor). The Compounding Debt Loop visualization would work well in client presentations about why investing in dbt modeling pays off.