The 10 ultimate tips for Data Governance
#5 Or how to impress your colleagues
Hello 😀
That’s it, switching to English 🚀
I had to give you a summary of the ultimate advice I've gathered from my experiences and numerous chats (or drinks) with you.
And it has to be said that I spend a lot of time reading articles and conversation threads on the communities I'm part of!
By the way, here's a list of my favorite resources.
Let’s go👇
Yes, you'll make mistakes...
It reminds me of when I launched my start-up.
No matter how much I'd read about Product Market fit, how to iterate very quickly on your product, how to find your very first customers, it didn't stop me from failing.
Launching and implementing a Data Governance programme is a long and difficult process, and you're bound to make mistakes along the way, such as spending time on details, fighting losing battles, and so on.
How can you optimise your Data Governance work?
But you can still avoid the biggest pitfalls by following these advice and logic.
1. Understand that this is a human challenge, not a technical one
It's one thing for you to understand it, but it's quite another for your boss and the members of Excom to understand it. Not to mention the IT department, who will no doubt tell you that they have already set up a data catalog, so there's nothing more to do.
Why is it primarily a human problem?
Because you have to align divergent interests to get there. And that takes time, because you have to convince and coordinate several stakeholders.
🔍 Tip #1 : Draw on feedback from similar companies: "At our competitor X, it took them 8 months to correct the quality problems with customer data that were costing them around €2 million a year, and the tool they used was deployed in just 3 weeks".
2. Get buy-in from key stakeholders
To do this, you need to speak the language of profitability. The idea is to build a business case with a sufficiently interesting potential ROI.
👉 I've already talked about this in episode #3 of this newsletter, with a concrete example to help you.
3. Start small but visible
If consultants tell you that you need to launch 10 domains in parallel, appoint 40 data stewards and make an inventory of all the data sources: run away!
So what's small and visible?
Take the latest data project or transformation project - is everything running smoothly?
Chances are it's not, because there's bound to be a small issue that's causing you pain: quality rules that aren't clearly defined, KPIs that are calculated differently depending on the source, and so on.
✅ This will be the perfect way to get started and prove yourself, without needing an insane budget.
4. It is not a project
So there's no end to it.
It's a bit like when you decide to take up sport, it's a lifestyle you want to install, not a fad that's going to last 3 weeks - even if that's the case for many of us 😅
It's a culture and a state of mind that you want to put in place in the organisation over the long term.
🔍 Tip #2 : Make sure you still break down your projects into small streams and set milestones to celebrate your successes.
5. Don't hire, but define roles
I don't understand companies that recruit Data Stewards.
➡ Data Stewards already exist in your organisation.
Do you think that sales people use dashboards with incorrect figures and carry on as if nothing had happened? ❌ No!
They'll do a bit of tidying up in Excel, ask an analyst they know, etc. And that's how they themselves clarify the data quality rules that work from a business perspective.
🔍 Tip #3 : Draw up role sheets with clear tasks and objectives, and get managers and the HR department involved so that these roles can be promoted through attractive career paths.
6. The Data Governance team is not the police
Or some sort of troublemaker who you don't want to invite to meetings. When, on the contrary, you should be acting across the board and providing solutions!
✅ How do you go about it? Bringing tangible evidence of what has worked for other subjects or other teams will give you instant credibility.
And if you've run out of ideas, Airbnb's story on Data Quality is a great inspiration.
7. Invest heavily in change management at a very early stage
Nothing is more effective than an email or video from at least the Head of Data, or better still the CEO. And it has to happen fairly early on in the launch of the program, so as to avoid the "I wasn't aware" or "No, sorry, I don't have the time for that".
Then you can scale up through a "Data 101" program explaining that everyone has a role to play in ensuring good data quality.
🔍 Tip #4 : Start simple: short videos, flash playbooks, articles posted on Teams, etc. All of this can clearly be done using existing free online content.
8. Leading by example is the key to data quality
Team managers must be exemplary in their own handling of data in the broadest sense :
By making documents available in the right place, with the right names and the latest versions
By proposing processes to alert on data problems
Taking an active part in resolving these problems
Communicating the importance of data to the business
In this way, they will inspire their team to follow their example, fostering a corporate culture focused on data quality.
😈 Because we all know very well that if your boss doesn't do it, you won't either...
9. Start with pragmatic tools
I can't say it enough, but...
❌ Stop chasing tools that are war machines when you don't even have a plan or roles in place !
To start with you can use simple tools that already exist in your company : a shared Excel to list domains and definitions, PowerBI and Python code for quality alerts, Notion or Confluence for documenting data flows, etc.
🔍 Tip #5 : Put the brakes on anyone who wants to start by spending 1 year putting out a request for quotation for a Data Catalog.
10. Put data in the hands of Business
At the end of the day, we can define processes, roles and rules, but the data is there to improve business decisions. We're doing all of this for Business teams !
It's up to them to define quality.
To take ownership of data-related problems.
To get involved in solving them.
Your job is just to make them want to do it.
🔍 Tip #6 : Find your first business ambassador and make him or her shine !
Want more?
👉 I regularly run Data Governance training courses in Paris at Hymaïa, there are still a few places left for the session on 30 May :)
See you soon,
Charlotte
I'm Charlotte Ledoux, freelance in Data Governance and AI.
You can follow me on Linkedin !