Hello 😀
I see Data Governance teams having their budgets cut and it breaks my heart.
Let’s see how to get money👇
Agenda
Impact estimation
Existing and new drivers
Tangible benefits
Impact estimation
2 visions have opposed each other since the dawn of time to estimate the impact of Data Governance initiatives.
Because, yes, that’s the biggest problem you’ll face.
Executives will challenge you. And they will do it at least every year when it comes to budget planification.
👉 So how can you measure Data Governance benefits? And make sure you keep your budget !
The Revenues-based approach
💬 You will get more competitive, you will get more customers, you will get more money. Sounds appealing right?
But, how do you link these gains with Data Governance initiatives?
That’s the tricky part.
Explain that : people will easily find available data, thus generating better insights and finally taking better business decisions leading to more revenues.
That’s the magic of linking Corporate Strategy with Data Governance Strategy :
In this approach, your revenues as Data Governance team need to be calculated through business KPIs, supported by the Data Strategy.
The Costs-based approach
💬 You will get penalties, you will endure high risks.
It’s like when we say that it will be 2 degrees hotter if we don’t reduce our CO2 emissions. The human brain is not made to take short term decisions for long-term risks. This is working well in risk-averse companies like insurance, banking or healthcare.
💬 You will lose opportunities, you will spend hours on data cleaning.
This one can be estimated based on how long it takes to deliver projects.
But we can be more precise than that. We can focus on time spent by Business teams. Why? Because even if you delete some tasks they have they won’t get fired, they will just be able to focus more on business tasks.
Existing and new drivers
To pick the estimation approach adapted to your context, you need to figure out your drivers. Too many companies don’t have clear drivers. They are doing Data Governance because the boss heard the concept at a conference.
I see 3 drivers for Data Governance :
Regulation compliance ➡ Costs-based approach
You have to respect a new data privacy or industry-specific regulation. Not doing data protection and policies would expose the company to expensive fines.
Business decisions ➡ Revenues-based approach
You want to improve customer satisfaction, decrease the cost of acquisition, predict your sales or improve any type of business decisions. Data Governance will help in providing high quality data for the linked use cases.
Operational efficiency ➡ Costs-based approach
You want to limit the time spent on data searching and data cleaning. By improving data quality processes, discovery and documentation.
What else?
There are new kids in town !
The AI Act ➡ Costs-based approach
Here Data Governance will help to perform risks assessment of algorithms and manage policies on training data.
Generative AI ➡ Revenues-based approach
Your data assessment service can provide high quality input data. This is a new challenge as it is unstructured data that is required. You could also facilitate ethical use through guidelines.
The product approach ➡ Revenues-based approach
Companies are now treating data “as a product” by offering it under various forms (dashboards, apps, etc) to Business teams. This trend requires several elements of Data Governance : accessibility to data for quick iterations, clear end-to-end ownership and interoperability policies between data products and domains.
Tangible benefits
Once your drivers are settled, you can start looking for tangible benefits.
Take your microphone, prepare your questions, let’s go !
Pick a major business team of your organization. Perform at least 10 interviews with operational team members, forget about managers and directors.
Data tasks interviews
A data task is any activity or operation that involves the collection, management, analysis, or utilization of data. It can be aggregation of data, cleaning in Excel, uploading a file to a system, etc.
The goal of these interviews is to estimate the data workload of people.
Here are some questions you could ask :
What is the data task you are performing?
What is the business purpose of the data task?
How often do you do it? (e.g. "Annually, Monthly, Daily")?
How many hours does it take approximately?
To what data domain do you think the data belongs?
Explain the step-by-step data flow from source to destination and specify whether it is automated or manual.
With these super simple interviews, you’ll catch what people do with data every day. You might be surprised, trust me !
Put the answers in a PowerBI, and create a heatmap to highlight which departments have the highest workload and on which data scope :
Perform these interviews before and after your first Data Governance initiatives.
And shine like a diamond by announcing tangible benefits :
We reduced by 20% the workload of Finance team on data manipulation
The delay of campaign creation is less than 1 month thanks to naming standardization
We decreased by 30% the time spent on collection of media data
But wait, these benefits are all on the costs-based approach?
Yes. Because you can easily assign a value to time spent.
It’s harder to assign a value to a new customer, or to an increase of customer satisfaction. And it’s even harder to define the contribution of Data Governance to these business KPIs.
Exit the revenues-based approach for now.
See you soon,
Charlotte
I'm Charlotte Ledoux, freelance in Data & AI Governance.
You can follow me on Linkedin !