Successful data management starts with master data

For many people, getting a handle on dirty data is an everyday occurrence, and they sit around every day mapping, adjusting and compensating for the fact that the data is dirty. In this blog post series, I focus on what data management is and why data management is a better alternative than fixing your data challenges with patchwork solutions. In this blog post, I also talk about what data you should start getting to grips with and why.

You've probably heard of transactional data, reference data, metadata, master data, sensor data, big data, SoMe data, etc. It can feel like a bit of a jungle, and you might be tempted to ask if data isn't just data and therefore think that you can just start from one end when you want to get started with your data management project. It is not quite like that. There are some core data that it is crucial to get a handle on first, because many other data can only provide value once the core data is in place. The core data is known as master data. Let me explain what master data is and why it is central.

Sometimes I like to say that master data, is data about the organisation's nouns. That is, it is data about, for example, Customer, Supplier, Product, Employee, etc. Referring to it as "nouns" often makes the right associations, but there are other data that are nouns too, so think of master data as the static data that is used over and over again and is a prerequisite for making a purchase, selling and delivering a product, paying an employee's salary, etc.

Master data is central to all organisations because it is what all our systems, processes and departments use to do their jobs over and over again. Take a product, for example. The purchasing department needs data about the product every time they want to order it, the goods receiver needs information about the product to receive it, the planner needs information about the product to plan, marketing and sales to sell it, logistics to distribute it, and finance reports on the product etc... I.e. some master data, such as product master data is used by virtually everyone in the organisation. If a piece of information is missing or a piece of information is wrong, then someone in the value chain is not happy. Imagine that there is a standard weight instead of the actual weight of the product. The logistics department gets tired of this pretty quickly because they get hit by it every time, they have to order freight.

Another example is if you, for example, have a lot of data about your customers' behaviour via orders, returns, customer support, payments, etc. Then you need to be in control of your master data to get anything useful out of it. If you have a lot of duplicates or wrong, flawed information about your customers, then you could end up with your conclusions about what your customers do or want being wrong. This can have consequences for your customer service and the decisions you make.

The good thing about master data is that it doesn't change very often, and that's fortunate because when it is so central and used by everyone in the business, challenges often arise when data suddenly differs across systems. So what to expect? Is it one or the other that is right? What was last updated and where? Sometimes there are problems identifying whether it is the same Customer, Product, Supplier you are talking about. People sit in their silos and desperately agree with themselves that it is best that they create their own system so that they can keep track of what is what. This is far from optimal, because what makes our processes, reporting and analysis work is that we use the same master data, so we don't end up talking about pears and bananas just because we've looked in different places.

While the above illustrates classic and key challenges of master data, it is also these types of issues you need to listen for when choosing which master data you need to get control of. To be successful in data management, it is important that you create real value and make a difference to those who need the data to be good. No one wants to do data management for data management's sake. Therefore, you need to start with the master data that is either painful today or where the risk of pain is greatest. That will typically be on the master data that goes across systems, functions, processes, etc.

In this blog post, I have addressed the fact that there are many different types of data, and it can therefore be difficult to see where to start your data management journey. While there can be plenty of challenges with other types of data, I recommend getting your master data right first, because master data is critical to the value of all other data, and because poor, missing master data challenges processes, decisions, and reporting. Finally, I recommend starting with the master data that cuts across all data, because it's typically the data that presents the biggest challenges or risks, and it's important that you use data management to create real value for your organisation. If the data management initiative does not create value, you will soon lose the goodwill of the organisation and the desire to collaborate and this can be fatal for your data management initiative. 

In my next blog post I will talk about how we need to work with governance to get our cross-functional master data under control.