Data quality: observability vs testing

Data quality: observability vs testing

Four reads on catching bad data before a dashboard lies to the board.

1️⃣ dbt tests vs Great Expectations — dbt tests catch what you anticipated; GE is heavier but expresses richer constraints. Source: Locally Optimistic.
2️⃣ Monte Carlo and the anomaly angle — observability watches freshness and volume for problems you didn't think to test. Via Monte Carlo blog.
3️⃣ Contracts over alerts — the shift to declaring schema expectations upstream so breakage fails loudly at the source. Source: dbt Labs blog.
4️⃣ Testing is cheap; observability is insurance — when each is worth the spend, by team size. Via Modern Data Stack newsletter.

Skip if your pipelines are five models deep. Essential if a silent null once cost you a campaign.

That's the stack for this week. Forward to a teammate.