Q&A: Transform Company Culture by Transforming How Workers Approach Data
Creating a data-driven company involves changing how data is viewed and used, both by top management and by curious, trained office workers. Frank Bien, CEO of Looker, explains how leading firms have transformed company culture by helping employees use data more effectively.
- By Linda L. Briggs
- January 19, 2017
Frank Bien is the CEO of BI and analytics company Looker and the coauthor (with Tomasz Tunguz) of a new book, Winning with Data -- Transform Your Culture, Empower Your People, and Shape the Future. In it, Bien shares techniques, tools, and tactics used by leading firms that have disrupted strategies and transformed company cultures by leveraging data effectively.
In this interview, he talks with Upside about some of the strategies he and Tunguz discuss in their book, including steps to creating a data-driven company. "We talk in the book about the importance of not just teaching someone how to answer one question or better manage data," Bien says, "but teaching skills that change the whole organization's approach to data."
Upside: You have an extensive background in big data and technology, including your current position as CEO of Looker. Can you talk about that background and how it led to the ideas in Winning with Data?
Frank Bien: I came into the big data space fairly early. Even before Greenplum [a big data analytics company where Bien was VP of business development until early 2011], I was at another big data company, and what I saw was that people and companies were collecting data at a whole new scale. Then with Hadoop the barrier to data collection became even lower. You could collect anything, essentially.
However, the tools to handle all this were still evolving. They were still built on the idea that the database was slow and expensive, that you had to extract data or put it in a data engine, ideas like that. That's why people weren't getting success stories from their data. Even with all the hype around big data -- The Economist put out a whole issue of the magazine around big data -- people still weren't having that much success with it.
That bothered me -- that there was so much investment and so little real business value. I asked: with all this investment and this big revolution in data infrastructure, how could we as technology vendors, analytics vendors, and BI vendors actually start delivering on the promise? That's the idea behind Looker and the ideas in the book, which are Looker stories. It all just came together.
Your book cites some great examples -- Twilio, Facebook, Google's hiring practices -- of the disruptive power of data. What's at the core of those disruptions?
What I find to be the most interesting part of each story is when a company can fundamentally shift its culture using data. I don't mean just using data to answer trivial questions, but fundamentally, so that every new employee is taught to use data to make better decisions. I never thought I'd be using the "culture" word, but it's true -- data is driving a new kind of culture.
We saw this early with companies such as AvantCredit, which is disrupting financial services. They have every employee start by attending a multiday boot camp to learn all of the data resources that are available and how to interpret them. Employees are taught the value of being curious, not in just one area but throughout the organization.
We talk in the book about the importance of that -- not just teaching someone how to answer one question or how to better manage data, but teaching skills that change the whole organization's approach to data.
Those stories are by far the most interesting to me. When your employees are really using the data from sales, from marketing, from products, then they can tell you why they are changing features on the front end and what that's doing to customer engagement. They can start to mix and match all of those pieces together because they realize they are all related. I might better understand customer engagement, for example, by looking at referral marketing to study who is referring new customers.
When you get away from siloes of data, all of that starts to tie together. That's why I find those to be the most interesting stories. We're talking about whole companies shifting, even in how they describe themselves, as it becomes a lot less about conjecture and a lot more about having informed opinions.
Your book talks about a common theme -- the importance of organizationwide buy-in for long-term success with data. What are some ways to build that buy-in? Where do you start?
I think it starts at two levels, and they're extreme opposites. First of all, it has to start at the top. If the CEO isn't open to new ideas, that becomes a problem. It's the effect of the highest-paid person in the room -- it just filters down. The executive team has to empower the organization to bring data and to sway them. Google and Facebook are two great examples of that. Data is always going to lead there.
It also starts at the bottom, with recruiting. You have to build a recruiting process that looks for intellectual curiosity; that's one of the most successful traits in new employees. They have to want to find the cause. Find them, hire them, give them the tools, and give them the chance to go find the answers. I think you have to have people who are curious to make it work. If you mix curious people who will go seek out answers with executives who can be swayed, you'll really start to build that idea of a data culture.
Speaking of hiring practices -- the book discusses how using data in an organized, applied manner can help leaders make better hiring decisions. As CEO of Looker, I imagine you use those techniques in your own hiring practices?
We definitely do. We use it not just in the hiring process but also in the "ramp" process. We really want to understand the types of people we're trying to hire and which types of people work out best. As they start, we can measure against our hiring criteria and see whether they deliver.
You really have to look at both sides, especially with a company such as ours that's growing very quickly. You want to figure out who you think is going to be a good employee, but then continue as a company to measure those criteria.
Looking ahead 18 months to a few years, where do you see us heading in terms of how companies manage, work with, and use data?
My reply might be different from what others are talking about right now. I fundamentally believe data will be the application. You're starting to see it now. We live in a world in which data is consumed through dashboards, separate tools, BI tools, etc. If you think of how important it is for your employees to be operating with data to make decisions, you see that your data shouldn't be separate like that.
New companies are emerging that don't have anything to do with BI. They are each doing a specific thing, but as a group they are largely surfacing data to people at the right time.
What we'll start to see is data moving out of BI tools and moving into the context of how people are working. We're going to start to see a lot more instances in which data becomes the primary application where people work. You'll come to the dashboard occasionally, but really -- if you're an account manager at the company managing customers, say -- you may be working in a tool in which you are actually consuming data to understand what customers you should be engaging, the likelihood that they're going to churn, and so forth. It changes the whole environment in which you work.
It's starting to happen now. Companies are realizing that if they really want people making informed decisions and operating more independently, they need to arm them with data, so we're starting to see the emergence of data applications that people are actually operating within.
How does Looker fit into what we've talked about today?
Looker is an exploration and discovery platform for data. We're not just a visualization tool or something sitting on top. We are a new kind of application server in the data world. We use an interesting architecture that connects directly to big, fast databases -- not surprising given my background. We first explain the data -- we create the common vocabulary. Call it data modeling if you want, but it's really creating a common set of metrics that people can use. Then, we allow people to explore using that -- not just one workbook or one silo, but the entire data world. It's the difference between a document and Google.
At the end of the day, we allow companies to integrate data into all of their work processes. Think of the ability to have one system that services sales operations, one for the marketing team, and one for the product team -- all of them using data coming out of one large data world. With that kind of approach to their data, organizations can be much more efficient.