Platform Engineering: Lessons from the Rise and Fall of eBay
summary
eBay executes a platform program that doubles engineering productivity by removing bottlenecks: automated pipelines for ~98–99% of apps, reintroduced feature flags, and measurement that moves deployment frequency 2/month → 2–3/week and lead time 10 days → 2 days. Mobile apps go from monthly to weekly releases and sitewide upgrades move from yearly → monthly through automation. Despite these technical gains, entrenched centralized planning, product arbitrage by sellers, and a Westrum‑pathological culture block strategic change and keep business metrics flat.
AI-written summary · about
chapters (12)
- 0:00Introduction to eBay's path
- 1:02eBay's historical context
- 2:43Identifying engineering flaws
- 5:31Defining success metrics
- 7:05The velocity transformation
- 10:18Execution strategies
- 15:31Fostering culture and scaling
- 23:27Results and reflections
- 28:54Structural and strategic hurdles
- 35:33Technology dead-ends
- 38:24The impact of culture
- 42:51Conclusion and Q&A
related talks
48mAgility and Architecture Together – Enabling a Fast Flow of Change
29mLeading With The Platform: A Platform Engineering Approach To Tech Debt
24mSignificance of Platforms in Product Development
46mAgility and Architecture Together – Enabling a Fast Flow of Change
23mPlatform Engineering for Architects - Crafting Platforms as a Product
42mLarge-Scale Architecture: The Unreasonable Effectiveness of Simplicity
35mThe elements of scaling
19mRevolutionize Efficiency: Mastering Platform Integration