This is an interesting point! And actually what you expect from your data practitioners should inform if data contracts are valuable to your organisation.

For example:

• Do you expect the data to be correct?
• Are you planning to use the data to drive important decisions?
• Do you need to trust the data?
• Is the data used to build any data-driven products, including ML-based products?
• Are the customers of those products internal or external?
• What is the impact if those products are unreliable or incorrect?

There's probably loads more! But by answering questions like these, you can decide how important it is to treat your data assets with the same kind of discipline you treat your software assets.

Personally, I don't think there is much business value in data that is delivered quickly, but can't be trusted.