Tableau CTO on the 3 traits of a data-driven enterprise culture

Data-driven culture has to be cultivated and nurtured, which takes time, says Tableau CTO Andrew Beers. Interestingly, the secrets of fostering data-driven culture aren't that secret at all – you can find them all in the anthropology section of any university library.

Andrew Beers Tableau CTO

Just about every research report looking at the trend of big data analytics technology tends to make a similar observation to one degree or other: the biggest challenge enterprises face in adopting big data solutions isn’t technology, infrastructure or cost – it’s the corporate culture.

Obviously this will vary from one enterprise to the next – digital-native start-ups are likely to be more at home with big data than a 150-year-old firm whose digital journey is still stuck firmly in the Excel age. But the maxim holds that in order for enterprises to capitalize on the kinds of insights and benefits that big data can provide to whatever business they’re in, it’s not enough to buy an analytics software solution and/or hire a data scientist who can understand how it all works – they need to have the right ‘data-driven’ culture in place.

And as many enterprises are learning the hard way, that ain’t easy.

Data-driven culture has to be cultivated and nurtured, which takes time, says Tableau CTO Andrew Beers. “So we’ve been thinking about how we can help companies effectively nurture these data-driven cultures.”

Interestingly, the secrets of fostering data-driven culture aren’t that secret at all – you can find them all in the anthropology section of any university library. Anthropology is the study of cultures, and it turns out that three key traits of cultures that survive and thrive over time have their equivalents in data-driven culture: language, adaptability and sharing.

Speak my language

Beers explains that language is how members of a culture express ideas and goals and collaborate to solve problems. In the big data world, says Beers, we call that data literacy.

“It’s really about being able to read, write and talk about [data], that’s our common language,” he says. “It’s about teaching people how to interpret data and use data, how to match up the right kind of question with the right set of data that will support that question, but it’s also about teaching them how to understand when data can be misleading – things like confirmation bias or understanding that correlation is not the same as causation.”

Beers adds that while every department should be data-literate to some degree, that’s not the same thing as requiring everyone to be data scientists.

“I don’t think most people have to become data scientists, but they do have to understand the right way to use data,” he explains.

Flexibility and governance

The second trait of successful cultures is flexibility and the ability to adapt to the changing needs of people in the group. In the context of analytics specifically, says Beers, successful flexibility lies in striking the right balance between the empowerment that flexibility brings with the necessary governance to ensure that flexibility benefits everyone in the group.

In a data-driven culture, he explains, “governance starts with providing a really, really rich experience around the community that comes together to work with their data … then we can attract them to an experience that has enough governance to allow that community to thrive. So think about governance as being able to trust the data in the system, knowing that you can get additional training and certification, and then you get insights on how all the data in your system is being used.”

Share and enjoy

The third anthropological trait of successful cultures is that the culture is shared with everyone in the group.

“You want to make sure you’ve got a community that embraces the data culture as a group, and then shares that and helps each other – helping them upgrade their skills, sharing best practices and things like that,” Beers says.

The key word there is “community”, which is also a core component of Tableau’s own business model. The company’s software is designed to designed to allow employees to noodle around with it, discover what they can learn with it – and then make the results go viral, either within company departments, across the company or even outside of their company via community-based apps like Tableau Public.

“That community does just amazing work – it’s everything from people learning courses to people managing their kids’ soccer league schedules, to people doing investigative journalism,” Beers says. “That large community is collectively teaching each other how to do good data analysis, how to tell a good story, how to put together a really good visualization.”

Beers adds that the Tableau Public community has done more to educate users on the possibilities of big data than the company could have achieved by designing tutorials for it.

That’s also why it’s important for enterprises to establish their own internal communities to explore big data, find innovative ways to use it and then share their experience with the rest of the company.

The importance of community

Indeed, Tableau reckons community is arguably the most important aspect of data culture precisely because big data analytics really delivers on its promise when you input more data into the engine – yet so many enterprises are structured so that data is compartmentalized in silos that have become so territorial that managers are overprotective of the data they’re sitting on.

“Communities tend to learn from each other and reinforce each other a lot, much more so than we can get by just teaching from the top,” Beers says, citing an example of an unnamed US enterprise customer whose IT group and business group were so siloed they didn’t even talk to each other.

“We spent an hour just convincing them of the need to work together, and over the next couple of years, they actually took our advice,” he says.

Two years later, the company would stage regular in-house mini-conferences where different groups would report their progress in becoming data-driven, show Tableau visualizations they had produced and the business value they achieved from them.

“That was great from a show and tell perspective. But then the department across the room from them said, ‘We need to talk with you later’, because they wanted to learn something from that group’s success, and apply it to their own group. And you only get that you have a successful community.”