I'm a fan of the sushi principle: raw data is better than cooked data.

Each process should take data from a golden source and not a pre-aggregated or overly normalized non-authorative source.

Sometimes the role of your system is to be the authoritative source of data that it has aggregated, validated, and canonicalized.

	▲	abirch 6 hours ago \| parent [-]
		This is great. Then I would consider the aggreated, validated, and canonicalized source as a Golden Source. Where I've seen issues is that someone starts to query from a nonauthoritative source because they know about it, instead of going upstream to a proper source.