Artiklar
Anteckningar om praktisk data governance
Korta, praktiska texter för data stewards. Hur du hittar datan som spelar roll, kommer överens om vad den betyder, och gör det till arbete du kan stå för.
/5 min läsningen
Write definitions that survive pressure
A definition dies when it is built to be filed, not used. The three parts that make one survive its first real argument.
/4 min läsningen
Why AI did not fix your data governance
You asked an AI chat to help write a data quality rule. The answer was good. A year later, governance is where it was. Here is the line between a sharp chat and a governance system.
/5 min läsningen
Cut your critical data element list in half, and defend every field that's left
A steward's critical list of around 60 fields got cut to just around 20 in an afternoon. The story, and the one question that did it.
/5 min läsningen
Build data quality rules the business actually trusts
Completeness 99 percent, and nobody trusts the data anyway. Why dimension-first rules fail, and what earns trust instead.
/4 min läsningen
End the meeting where the numbers never match
The same report, two numbers, every quarter. The meeting repeats because the fix was wrong, not the data. One question tells you which problem you actually have.
/3 min läsningen
Make data governance that people actually use
If governance feels like theater, you are doing the version designed to feel that way. The one shift that makes it real.
/3 min läsningen
Say no to the data requests that aren't worth your time
Most data requests are not worth the work. The questions and scripts to push back without being the difficult one.