37% of data incidents are caused by schema changes

Barr Moses, Co-Founder & CEO at Monte Carlo, posted some interesting data on data incidents on LinkedIn.

According to data from their product, the industry standard is 6 data incidents per 1,000 tables per month. Of those:

• 37% are caused by schema changes
• 28% are caused by freshness (presumably, lack of)
• 23% are caused by changes in data volume
• 12% are caused by quality issues

This is just one data point, but anecdotally it doesn’t feel too far off.

Certainly unexpected breaking schema changes are a big problem for many data teams.

And that was the first problem I solved with data contracts.