Have you ever been in a meeting with a data scientist and you look around the room to see most of your coworkers with bored looks on their faces? The scientist goes on and on, on a dry speech about numbers and their implications and, by the end, it seemed like the room barely grasped what they were saying. Sounds familiar?
Storytelling is one of the most important skills for a data scientist. It doesn’t matter how ingenious your models are, if you can’t build a story that other people can understand then your data isn’t doing anything to change the world.
Having said that, the burden of understanding how data works shouldn’t fall on the data scientist alone. Yes, you have the responsibility to build stories and to choose the right data visualization strategy. Still, one can only go so far when other people don’t understand even the basics of data analysis.
Beyond The Indicator
Did you know that there is a mathematical correlation between the rise of temperature in the Caribbean seas and the reduction of piracy in the area? It may sound like a joke, but it’s not, the correlation is there.
That’s what it’s traditionally called a spurious correlation. It’s what happens when you can establish a mathematical correlation between 2 or more indicators, but that correlation doesn’t make sense in the real world.
Linking pirates and climate change, funny as it sounds, is an extreme example of something that happens all too often—taking numbers at face value. Long gone are the days when an executive summary was a single-page report with a single indicator and a recommendation. We live in a data culture, and you have to adapt.
Data culture is the collective set of behaviors, beliefs, and traditions of people who value, practice, and encourage the use of data to improve decision-making. In data-driven environments, every member of an organization understands the importance of data and bases their decision-making on empirical evidence.
Data-driven environments stand in stark contrast to traditional strategic decision-making. It trades experience and intuition for a more empirical approach where decisions are strongly supported by evidence and data-based projections.
A (Data) Cultural Revolution
Studies have shown that companies that embrace a data-driven culture tend to show growth in both output and productivity between 5% and 7% in the short term. This translates to annual growth of up to 30%.
Despite its many benefits, embracing data culture has been an uphill battle. In a recent survey, over 74% of managers reported that they haven’t been able to implement a data-driven culture in their businesses. Ironically, the same number of managers reported that they fear business disruption by data-driven competition.
In other words, we understand the importance of data, but we haven’t managed to find the right strategy to push towards widespread data-driven adoption.
One of the biggest problems with data culture is convincing people of the importance of data gathering and analytics. To implement data-driven decision-making, everyone involved has to value the benefits of making strategic choices based on data.
Playing the devil’s advocate, resistance and mistrust towards statistics isn’t unwarranted. For decades, we’ve seen people abuse linear models, make rampant predictions with small sample sizes, and misconstrue the implications of their results.
But we are no longer in the 20th century. We have more refined statistical models, data mining, and the support of AI and machine learning algorithms. We have more processing power, which leads to better and more accurate predictions.
Introducing Data Literacy
Data literacy is the ability to read, write, and communicate data. This includes an understanding of data sources, analytical methods, constructs, data cleaning, and statistical models.
To take it one step beyond, it also encompasses the ability to describe use cases and applications, as well as to build stories from data.
Data literacy, like any skill, can be learned, and it comes in different levels of expertise. Not everyone in your company needs a Ph.D. in statistics to be data literate. It’s enough to understand the basics, to know how to ask the right questions, and to know where to look for answers.
Data literacy promotes decision-making by creating an environment where all members of an organization can ask the right questions about the data they are working with. They have a clear understanding of the limits of mathematical models and can detect anomalies and other limitations.
This makes you more cautious towards the conclusion you can draw from data and, at the same time, it helps you understand the ways how data can be analyzed and interpreted.
To promote data literacy you have to promote data culture first—and that’s something that isn’t going to happen overnight. Slowly, but surely, you need to ease a transition, first by creating a positive attitude towards data, and then by motivating your team to learn about it.
Tips To Build a Data-Driven Culture
First and foremost, we wouldn’t be here if not for human creativity, instinct, and imagination. The “human element” is something that can’t be replaced. Reassure your team, and let them know that their gut feelings aren’t going anywhere, but rather, that you’re giving them more tools to make better decisions.
That might seem like a very specific piece of advice. But the underlying principle is simple: people, especially in groups, have a hard time accepting and adopting change, mainly because they may feel threatened by it.
As new positions open up in your business, look for candidates that show signs of data literacy. Once again, they don’t need a Ph.D.—it’s just enough that they understand how data works. Ask questions during interviews that explore their knowledge and understanding of data analysis.
Once data-literate people join your team, you can help them coach other members who are having a tough time. Some companies even create data-focused offices whose sole purpose is to promote the use of data in all areas of the business.
Naturally, changing people’s perspectives won’t amount to much unless you also start adding data-centric activities in your workflow. The more you show how much data can help in your business, the more people will feel motivated to adopt a data-driven culture.
Last, but perhaps not least, be ready to invest in the right kind of technology for a data-driven culture. Develop or buy software that facilitates the data gathering and analysis process. There are dozens of tools out there that go as far as analyzing and interpreting the data with very little input from the user.
The rest is basic advice: create workshops and offer incentives to become data literate. As soon as you start seeing the benefits of creating a data-driven culture, share the results with the whole company. Let people see firsthand what they have to gain by becoming data literates.
The process is far from simple, but more and more companies are getting on the data-driven bandwagon. Businesses who stick to traditional methods are still safe, but as time goes on, the chasm will widen and data-driven companies will have the competitive edge. It’s time to change.
