How to sort out Data Quality and make your CEO love you

If you want better data quality then don't start by improving data quality.

• Cleaned up data feels rewarding and you can show improvements.

• But did you consider whether the cleaned data helps anything?

• Did you consider how to stop the errors from recurring?

• Why was it entered incorrectly in the first place?

Start with these three actions

Discover which data items have the most impact

Talk to a finance person and they'll tell you where faulty data has a significant impact on the year-end accounts.

Talk to a sales person and they'll tell you where faulty data affects their conversations with prospects and customers.

Talk to a distribution person and they’ll tell you where faulty data causes deliveries to fail.

So why guess? Get out there and ask people.

Find the break point

Maybe you are lucky enough to be in an organisation with well-documented Data Lineage. In that case you can track back from the consumer of the data and identify where the error was introduced.

For the rest of us, you need to follow the trail. The end users will be able to tell you which report or dashboard they use, and then it's up to you to do the legwork to find out what data fed that report, and how it got there. Sometimes, the creator of the report has left the firm and didn't leave detailed documentation, but you should still be able to find out where the data comes from.

If that report is the cause of the error, and you don't have documentation, there could be some detailed unpicking to do. It might be easier to rebuild the report itself!

Understand the cause

Occam's razor is a principle that simple explanations are more likely to be correct than complicated ones. If data is entered in a strange way, or goes through a nest of transformations, that complexity might not be necessary. More likely it is the result of multiple processes which have adapted over time.

Don't be afraid to ask why things are done as they are.

Certainly don't be afraid to ask if things can be done differently, but remember to explain who this helps (not just what and why) down the line.

Fixing the root cause of data issues means less rework, more trust, and better decision making.

What’s your biggest data quality challenge?  Tell me and I will send you back THREE actions that YOU can take to help.

Have a wonderful week,
Charles

P.S. Was there anything in the newsletter that might be useful or interesting to a friend or colleague? Why not share it with them?

P.P.S. Do you have an article or ideas that you want to share? Send me an outline to be considered for a guest posting slot.