↓Skip to main content

The language of data contracts

·2 mins
Data Contracts

Diagram showing enforcement not getting engagement from a person, and enablement does get engagement from a different person.
The language we use is important.

It helps define culture. It shapes behaviours. It defines our identity.

Which is why I find the language used with data contracts interesting.

A lot of people talk about how they can use data contracts to enforce something on their data producers.

Enforce better data quality. Enforce data governance. Enforce change management.

And while data contracts can be used to enforce those, using that kind of language is unlikely to encourage your data consumers to use them.

You’re already starting off on the wrong foot, implying blame, and creating friction.

When what you’re aiming for is understanding, collaboration, and alignment of incentives, so that quality data is produced at the source, for the benefit of the organisation.

And data contracts enable your data producers to do easily provide that quality data.

So, before talking about data contracts to data producers, think carefully about the language you’’re going to use.

The End of the Bronze Age: Rethinking the Medallion Architecture by Adam Bellemare

I’ve never been a fan of the medallion architecture and the assumption that data engineering teams need to accept raw data (bronze) and do all the work to make it gold. In fact, someone asked me once if by moving to data contracts we’re aiming straight for gold, and I think generally we should be.

The Rise of Single-Node Processing: Challenging the Distributed-First Mindset by Alireza Sadeghi

A good, detailed explanation on why it may be worth moving away from big data systems towards singe-node data systems such as DuckDB.

Is engineering strategy useful? by Will Larson

Substitute Engineering for Data and everything still applies.

Being punny 😅

I’m not one to brag, but I finished a jigsaw puzzle in 4 days. The box said 3-5 years.

Upcoming workshops

Thanks! If you’d like to support my work…

Thanks for reading this weeks newsletter — always appreciated!

If you’d like to support my work consider buying my book, Driving Data Quality with Data Contracts, or if you have it already please leave a review on Amazon.

Enjoy your weekend.

Andrew

↑