Working in small batches: overcoming cultural barriers
summary
Argues that reducing batch size cuts feedback loops, lowers sunk cost and risk, and unlocks delivery speed by shrinking transaction and wait time in pipelines. Uses CI/CD pipelines, feature toggles and pull‑request workflow changes as levers; measures batch size via deployment frequency and merge cadence. Empirical result inside Bloomberg: 40% reduction in PR size, 20% more PRs, and 50% faster merges after culture conversations and pipeline tweaks, demonstrating small batches work when transaction costs and queueing are addressed.
AI-written summary · about
chapters (10)
- 0:00Introduction and background
- 1:06Defining continuous delivery
- 3:58Culture vs technology challenges
- 5:01Why work in small batches
- 8:35Optimal batch size and costs
- 12:30Transaction costs in delivery pipelines
- 14:16Code review batch size impact
- 17:52Overcoming cultural barriers
- 20:43Practical steps to reduce batch size
- 21:07Q&A session
related talks
29mEconomies of Speed
22mKeeping builds green at scale
54mBridging the Gap CICD Mind the Gap
27mOptimizing CI/CD Pipelines for Faster Time to Market
29mHow thinking small is changing software development big time
46mContinuous Deployment Workflows
42mSpeeding Up Innovation
40mBetter Living by Changing Less – IncrativeOps