The DevOps Handbook has a framework called CALMS. Most data teams only do the A. Here’s what they’re missing.
CALMS stands for Culture, Automation, Lean, Measurement, and Sharing. It’s the DevOps Handbook’s framework for organizational improvement. Data teams know about it. They just fund one letter.
C - Culture. Producers don’t talk to consumers. Ownership is unclear. When a dashboard breaks, everyone points somewhere else. Culture eats data strategy for breakfast - and this is the letter nobody budgets for.
A - Automation. Here’s where most investment goes. Pipelines automated. Deployments manual. Testing skipped. Quality checks: none. Half-automated systems create false confidence. I’ve seen teams spend months automating ingestion and zero hours automating the tests that would catch a breaking schema change.
L - Lean. WIP limits for data teams. Stop starting new pipelines and finish the ones already broken. Every half-built pipeline is inventory that rots. Lean asks: what’s the cost of carrying all this unfinished work?
M - Measurement. Teams measure pipeline uptime and row counts. Nobody measures adoption, time-to-insight, or whether consumers trust the data. You’re measuring the factory, not the product.
S - Sharing. Documentation lives in one person’s head. Knowledge hoarding is the default, not the exception. Sharing is a practice you build into process - it doesn’t happen because people are generous.
The imbalance is consistent: 80% Automation, 5% Culture. Reverse that ratio and most data quality problems start solving themselves.
If you scored your data team on each letter right now - which one scores lowest?