Data Mesh in Practice

Chapter 3. Getting Started: A Data Product–Centered Mindset Shift

One of the biggest misconceptions when working toward a data-driven organization is that change can be forced by technology. When people hear about a new concept that is supposed to change all their data-related problems for the better, they look for the shortcut, the easy way out, the tool to apply that solves all their challenges. By now, most data practitioners understand that even a change of technology is often a mid- to long-term project, especially when it requires migration of hundreds to thousands of internal data use cases. Therefore, a fundamental change of direction cannot succeed with hurried short-term plans but needs a bigger shift of mind instead.

In Chapter 2, we introduced the idea of data products and outlined the advantages of having domain experts taking the ownership of data they are offering to others. To apply these ideas in practice, we have to go one step back and start with what is often the status quo. Let’s reflect on how data is commonly shared and processed across many organizations.

The Rise of Big Data Technologies

We often divide data between two different purposes, transactional and analytical. Transactional data is anything that supports the day-to-day business of your company. It can range from compara­tively raw data—like an order being placed in a shop—to complex messages that are exchanged for communication between two services. Analytical data is changing the focus toward ...