The term “data driven” has become a mantra of modern business management. Almost every business wants to be a data-driven enterprise. Yet most organizations haven’t even set forth a definition of what it means to be data driven. A simple web search turns up dozens of definitions—some related to decision processes, some related to metrics, and some related to business results such as quality of service. But none of these generic and somewhat academic definitions is about you and your company.  It is also important to distinguish between data-driven vs. data-informed, and to achieve the right balance of the two. The concept of data driven must become tangible and personal. Every data worker and every knowledge worker should understand how it affects them and how they influence it. Data-driven practices need to be woven into data culture.

What is Data Culture?

Data culture is the collective beliefs and behaviors that shape an organization’s capabilities to turn data into information, knowledge, and insights that drive decisions and actions to produce positive business results. 

In general, culture is always defined by people and what they view as behavioral norms and expectations. Long-standing and widely-accepted norms and expectations become group customs. These cultural generalities hold true for data culture. Data culture is defined by what people think, say, and do about data. Data customs are an integral part of data culture. Good data culture occurs when data customs include these behavioral norms:

• Leaders lead by example, openly demonstrating how they’re guided by data.

• Leaders trust people to use data in the right ways to guide decisions and actions.

• People trust the data that they use.

• Critical thinking, questioning, and learning from data are encouraged.

• Data literacy is valued, recognized, and incentivized.

• Data and the tools to analyze data are available to everyone. 

• Sharing of data and sharing of analysis are core values.

• Data is woven into business processes.

• Data capabilities range from basic reporting to advanced analytics and data science. 

• Business people ask good questions that can be answered with data.

• Data people answer those questions in a way the business understands and can act on.

• The distinction between business people and data people becomes insignificant as business people become data literate and data people become business literate.

Shaping Data Culture

Getting to the right data culture for your organization requires management. As Wayne Eckerson says in his article about Creating an Analytics Culture, “The data culture in many companies resembles the Wild West—everyone relies on their own data and resources and distrusts everyone else’s.” The challenge of managing data culture is that many traditional management styles don’t work. You can’t create culture through visioning, directive, democratic process, or other conventional management methods. In fact, you can’t create data culture because it already exists. 

If you have data, you have data culture.
What does it take to reshape existing culture into desired culture?

If you have data, you have data culture. Managing culture is a process of coaching to reshape existing culture into desired culture. Introduce desired data practices at a managed pace, then work to achieve the acceptance and durability that is needed for them to become data customs. When it is customary to inform decision processes with data, to apply critical thinking, to pursue data literacy, to share data and analysis, and to practice other desired behaviors with data, then you have managed data culture.

Culture, behaviors, practices, customs—these are all big and abstract things that are difficult to tackle with coaching techniques. We really need to break it down and make it a bit more concrete. Figure 1 shows five dimensions of data culture that make practical targets for data coaching. 

Figure 1. Dimensions of Data Culture

Mindset addresses what people think relative to data—the attitudes that they have that shape (often subconsciously) their behaviors when working with data. Coaching goals for mindset include guiding people to value data, seek facts, and spontaneously analyze data.

Skills cover the range of capabilities that people have for working with data. Some of the important skills include data literacy, analysis capabilities, and data storytelling.

Work style is a significant part of data culture because culture is really all about group behaviors and norms, not individual actions. A culture of collaboration, communication, and shared insights moves you well down the path to data driven. When data and analysis drive conversations, then you know you’re on the right path.

Tools are fundamental to working with data, and the right toolset can actually be a lever to help shape data culture. Among the toolset goals consider integration and interoperability, self-service, and an enterprise data catalog. The catalog is an enabler of knowledge sharing and collaboration.

Data is, of course, a key part of data culture. Data that is governed, high-quality, and trustworthy is critical to engaging people and having them work confidently with data. 

Each of these dimensions of culture has a role in becoming a truly data-driven organization. For each I’ve given only a few examples of the coaching goals. There are certain to be many more, and you’ll want to align them with the norms and customs that you identify to define and describe your desired data culture. 

Getting to Data Driven

Data culture is an important ingredient, but culture alone does not make a data-driven organization. Data leadership is equally important, and data leaders fill critical roles in creating and sustaining a data-driven organization—they are the data drivers.

Nothing will be driven unless you have drivers.

The harsh reality is that nothing will be driven unless you have drivers. To be data driven, you must have data drivers. They are the data leaders that you’ll find among data owners, data stewards, skilled data analysts, highly data literate business analysts, compelling data storytellers, etc. Find these people and know what kinds of data leadership they bring to the organization and the effort. They are pivotal when it comes to changing culture and becoming data driven.

To illustrate the importance of drivers, let me shift for a moment to some basic principles of mechanical engineering. Imagine two gears that turn with teeth interlocking. One gear has power applied to it from a motor or other source. It is the driving gear that causes the other gear (the driven gear) to turn. Without the driver, the gears don’t turn.

That’s pretty basic. But let’s go a little bit deeper. The size of the gears—both driver and driven—determines the speed at which things happen and the amount of force that is applied. (See figure 2.) A small driver applied to a large driven gear increases force but reduces speed. The gear with 8 teeth driving the gear with 16 teeth results in ½ speed and doubling of torque for the driven gear. A large driver applied to a small driven gear reduces force but increases speed. The gear with 16 teeth driving the one with 8 teeth results in twice the speed at half the torque for the driven gear. 

Figure 2. Driving vs. Driven

So, what does this mean for data-driven organizations? To become truly data driven, you need to change culture, behaviors, practices, and customs. To achieve the changes, you’ll need data drivers—your data leaders. If you’re making big changes, then expect to make them slowly with a small number of data drivers influencing a large population of data and knowledge workers. The alternative is to make lots of small changes quickly. When you have a large number of data leaders, they can drive change in small populations of data and knowledge workers quite quickly. 

These concepts tie directly to data strategy. One organization that I’ve worked with recognizes that they need to start by making some big changes with only a few committed data leaders, and that it won’t happen instantly. They expect that with those changes the population of data leaders will grow and they can evolve from big-and-slow to small-and-quick. The long-term vision is what they think of as agile data culture—recognizing that the culture will need to change as data and business dynamic change.

Final Thoughts

Becoming data-driven isn’t easy but it can be valuable. Managing data culture isn’t easy but I’ve come to believe that it is essential. If you manage data culture well, data-driven will happen. If you blindly pursue data-driven, data culture may suffer irreparable harm.