You're frustrated.
You're frustrated that the data's not right.
You're frustrated because the data comes from multiple sources which don't join up.
You're frustrated because different people have different numbers for what should be the same thing.
Your data is a mess.
No one wants to do anything about it, because the business is prioritising other activities.
The rebrand and the cool new logo.
The licence fees for that tool which "solves Data Governance".
The training courses which map everyone to a particular colour.
But you want to sort out the data and you haven't got the resources to do it.
Then you found this article and thought that, maybe, it's worth a read.
Isolating one activity from a data governance framework is risky, and I expect some debate in the comments. I've even changed my own mind whilst planning this article.
Because it's obvious that the one to choose should be about people.
People first.
Culture.
Attitudes.
Responsibility.
Then I changed my mind.
After all, what is the point of engaged people if they don't know what to do and aren't empowered to do it?
That just creates disengaged people.
And then your next attempt at improving data just got harder because the last one didn't work.
So, Charles, what is this single most useful activity?
It's the Data Quality Log.
Why?
Because people are fixing data all the time, but the fixes are "one off".
Whereas resolving data issues at source, resolves them sustainably.
And data problems that don't get created; don't cost anything to fix.
A Data Quality Log records all the data issues which come up, who is impacted and how, what you are going to do about it, why and when.
There's a few more columns that you need to have (e.g. An ID so you can track them) but it's fundamentally about the what, why, who, where, when and how.
You can do it with a spreadsheet. I'll even share a template if you ask.
When should you use your Data Quality Log?
Use it whenever a data issue crops up.
Don't be tempted to skip the "easy ones" or the ones you fixed. You want them all so you can spot the trends.
Escalate when you need to. Some items may need stakeholder co-operation to resolve; or for the right person to accept the risk of leaving an issue unresolved. The biggest or longest standing issues need to be escalated further so that people are aware of them.
There is lots more you can do to improve your data, but this one activity will get you going.
Over to you
- Share your experiences in the comments. What has worked for you; and what has not gone so well?
- These articles get published first at datazed.co.uk/writing, but I send out weekly emails with information about upcoming events and actionable advice, sign up at datazed.co.uk/register
- Join me for my next Data Governance Get Together. Some of them are in person, some are online and some are both. Find out when the next one is at datazed.co.uk/events
- Work with me 1-1 to develop a Data Governance plan for you and your business.