My Top 10 Data Governance Best Practices

#34 Share to Top Management

Hello 😀

🚨 Alert : the newly Data Governance Council appointed asked you to give them some Data Governance best practices…

Nice hike to the Rocher de la Dame Jouanne, giving full view on Fontainebleau’s forest !

• Join my next online Data Governance Training Day

• Play the Chief Data Officer simulation game

• Get my Data Governance templates

• Discover my consulting services

• Join +280 readers of “Data Governance : where to start?”

• Get a boost with a 4 weeks training on Generative AI

---

📢 The Chief Data Officer game is brand new !

We’re partnering with Informatica, and you're going to love what’s new in the game : 💎 4 special treasures to find through all emails, new emails covering Data Quality, Data Governance, Golden Data, AI Best Practices… and 🏆 a unique opportunity to participate in the Informatica CDO Academy !

---

Let’s see what to say 👇

1. Setup foundations

2. Operationalisation

3. Embedding & automation

---

Setup foundations

The Data Governance Council needs advice first on where to start and how to lay strong foundations. Here is what I always say first to the Top Management :

1️⃣ Align governance initiatives with business goals and organizational strategy

Clarify why Data Governance should exist in your company. How will it serve the business strategy? It should overall help to support better decisions, innovation and competitive advantage.

Keep in mind also that setting up Data Governance will have an impact on the organization.

🔍 Tip : If you’re planning a massive reorganization, it’s probably best to wait a bit before starting Data Governance.

2️⃣ Consider data governance as a long term investment not a one time project

It will take time, but the reward is worth it : your company will stay afloat by using its data wisely and effectively. The consequence of this is that the “central Data Governance team” is here to stay. Hence they need at least a super strong support from you (= Top Management).

If this is not acknowledged, no need to go further.

3️⃣ Invest massively in change management and communication

If you have to choose between a tool and a global training : pick the training! We always underestimate how much time and energy it takes to have people changing their habits. That’s what Data Governance is about. It’s about changing mindsets and the ways of working. 🤯 Which is so much more difficult than plugging a new tool.

👉 And you’ll have to operate these changes at all levels within the company: from the executives to operational team members. That’s a massive challenge.

---

Operationalisation

Now that buy-in is acquired, the framework is stabilized and validated by sponsors and key stakeholders. Let’s dive into making governance operational :

4️⃣ Prioritize quick wins and incremental improvements

You need to manage expectations. Be careful when you start receiving many new demands from business teams and/or many quality issues sent to your team. Prioritization is king. And more than that is having the prioritization result stamped by the Exec Committee.

🔍 Tip : Select small battles with high impact. The most critical KPI of the company is still not defined? Gather the 3-4 people most affected and get them to come up with a common definition.

5️⃣ Assign clear ownership and stewardship

Don’t hire outside of your company, but assign roles to business team members. Let them know that the tasks associated to the role are things that they actually already do now : clarifying a definition of a business concept, coming up with a calculation rule for a new metric, etc. They do it daily without thinking about it.

👉 What will change is how they’ll do these tasks, how often and with which tools.

🔍 Tip : By clear ownership, I mean having the name of a person, not a team’s name. That’s how people own stuff and become responsible. If the organization changes, the Data Governance team will make sure there’s a new name in the owner box.

6️⃣ Start with lightweight policies

I’ve done a list of the first policies to start. It’s basically around the standardization of your naming conventions, data formats and ownership. You should also start looking at access management by doing a global audit to understand risks and next steps.

Remember that these simple policies should live in the long term, hence you need a clear validator and owner of policies to have them respected and audited regurlarly.

7️⃣ Reuse existing tools

You don’t need AI-powered modern tools to start your Data Governance. You should reuse what you have :

• a shared digital workspace to store your policies,

• a Confluence to start documenting business concepts,

• a Notion table to list your critical KPIs,

• a dbt pipeline to do data quality checks, etc.

• even an Excel file can do the trick to initialize a data catalog !

---

Embedding & automation

Great, you’re now thriving and going on run mode for your data governance initiatives. What’s left is :

8️⃣ Embed governance into existing project and business processes

Please don’t treat Data Governance as a separate track. You have to integrate it into project lifecycles, workflows, and decision-making routines. This ensures governance becomes second nature for teams, and avoids bureaucracy.

🔍 Tip : Take your current project processes and add to them some simple governance tasks to each milestone like having a quality check performed during the framing phase, getting definitions of business terms involved in the data catalog before the development phase, etc.

9️⃣ Automate data governance work where possible

By automating recurring governance tasks like metadata capture, quality checks, or policy enforcement, you free up valuable time from your Data Stewards and reduce human error.

👉 Automation makes governance scalable.

🔍 Tip : You can start automation on small and simple tasks like the Data Stewardship inbox triage, with very simple tools. But you should also rely on your Data Engineers to automate the work.

🔟 Data Governance means animating, repeating, re-explaining

I’ve spent many hours explaining concepts, simplifying them for various audiences. The more I did it, the more people would start understanding and getting involved. Add to that the creation of a fun community around Data and you’ll see the organization changing.

👉 Don’t underestimate the time to animate a community in the long term though.

🔍 Tip : Find your close soldiers, aka Data Ambassadors, and have them re-explain things and being actors of change for local business teams.

See you soon,

Charlotte

I'm Charlotte Ledoux, freelance in Data & AI Governance.

You can follow me on Linkedin !