Information Week
9 Secrets of Data-Driven Companies
By: Lisa Morgan
What does it mean to call your company "data-driven?" Definitions range from simple reporting to viewing data science as a core business strategy. We asked executives from a variety of businesses to help us identify which traits are essential for becoming a truly data-driven company. See what we learned, and tell us how your organization stacks up.
The competitive landscape is shifting rapidly in virtually every industry, fueled by the intelligent application of technology and data. While nearly all companies have more data than they know what to do with, a powerful minority is discovering new ways of leveraging it.
"Everyone is trying to figure out what the best route is, and what value data science can bring to the business," said Vivian Zhang, CTO and founder of the NYC Data Science Academy.
Internal fragmentation is one of the biggest obstacles facing companies as they seek to become data-driven. Data remains trapped under the control of a given department or business unit. The fact often frustrates efforts to use data strategically across the enterprise.
"Many organizations have made progress in some area, but are still deficient in other areas. That speaks to the challenge of bringing everything together to optimize it and to take advantage of what the promise of big data holds," said Tim Herbert, senior VP of research and market intelligence at industry association CompTIA.
Of course, orchestrating everything requires more than technological problem-solving. It requires cultural transformation, which is often the largest stumbling block of all.
"To create an ecosystem in which data-driven culture thrives, you have to have a strong balance of strategy, IT, and statistics, all working in harmony," said Joshua Jones, managing partner at analytics and intelligence consulting firm StrategyWise. "Too often we see academically focused projects that lack a business output. Or, you know what your competitors are doing, but your IT systems aren't there. [Alternatively], you have great IT systems, but you don't have the statistical expertise that enables you to get beyond reports that show you in the past."
The term "data-driven" itself is being challenged. The phrase has been around for about a decade, and much has changed in that time span. Alternative descriptions include "data-informed," "insight-driven," and "data-science driven," all of which reflect a broad state of data-savviness. The first option focuses on decision-making, the second stresses action, and the third reflects use of the scientific method. The last concept is worth noting from a cultural perspective because the most advanced companies are less likely to confine scientific thinking to data scientists or the data team.
"A company becomes data-driven when it subscribes to three things," said Sham Mustafa, founder and CEO of data science job marketplace Correlation One. "One: Everything can be measured. Two: Data reduces uncertainty and drives better decisions. Three: Data is a part of the core business strategy."
The bottom line? The notion of becoming data-driven is continuing to evolve as the market matures. Here's a look at the characteristics of today's most sophisticated data-driven companies. Once you've reviewed these, tell us how your company stacks up in the comments section below.
Comfortable With Uncertainty
Benjamin Franklin said, "Nothing [is] certain except death and taxes." Yet, professionals without statistical backgrounds have historically been more comfortable with single-number answers than comparative levels of uncertainty. This is changing as companies begin to promote scientific method-type thinking throughout the ranks.
"Sophisticated companies have a culture of hypothesis-testing and experimentation. [They] basically look at every decision as a problem of choice of their uncertainty," said Sham Mustafa, founder and CEO of data science job marketplace Correlation One. "The spirit of the scientific method runs throughout the organization."
Adapt Culture Accordingly
Data has always been central to the core business strategy of companies such as Google, Facebook, Amazon, and Uber. Organizations have to think hard about how they can leverage data in new ways. They have to cultivate a culture of curiosity, creativity, experimentation, and collaboration.
"It starts at the top. There has to be an intellectual curiosity and curiosity about the business impact across all departments," said Ron Heinz, founder and managing director at VC firm Signal Peak Ventures, in an interview. "Less sophisticated companies that don't have the curiosity won't be as successful."
Mortgage marketplace Sindeo displays its key business performance metrics on TV screens throughout the office so that everyone in the company is aware of them. The company also has weekly meetings in which employees are encouraged to ask the executives and the data scientists anything. According to Sindeo Data Scientist Matt Cornett, at least two questions a week are data related.
Be Agile
Data-driven companies tend to be more agile than their less mature counterparts, which means they're willing to adapt the business as necessary to stay competitive. Reorienting a company's vision is one thing, executing that vision is another thing entirely.
"Becoming data-driven has to be part of the organizational DNA. The good companies ask themselves if data is part of their core strategy. Another question is whether they have the right organizational structure that supports their strategy," said Sham Mustafa of Correlation One.
Forward-Looking Technology Acquisitions
The last thing a data-driven company wants is barriers to information flow. While privacy, security and governance are important, so is the ability to use data in ways that provide new insights. Quite often, that involves combining different datasets, some of which may not have been previously available.
"Most organizations have a long history of duct-taping software packages together. It's important to understand how the different systems talk together or don't," said Joshua Jones of StrategyWise. "An advanced organization has taken those things into account and ties all the data points together."
Update Business Processes
Tools and technologies can only take a company so far. One common misstep is failing to adapt business processes so the company can effectively leverage its use of data.
"Sometimes, companies acquire tools, but they don't have a mandate around the implementation and making it successful," said Ron Heinz of Signal Peak Ventures. "Some businesses have been able to get away with outdated processes, and it doesn't matter as much. In other industries, such as software, if you don't have that implementation, you're going to be left behind."
Interestingly, some companies that consider themselves data-driven are less mature than they think, because, upon close inspection, they're still failing to align business processes with data-driven decision making.
CEOs Lead By Example
It's one thing to demand your people use data, and quite another to demonstrate data-driven decision- making, critical thinking, and curiosity. When the CEO and other business leaders lead by example, they are also more inclined to make the technology and human capital investments necessary to enable a data-driven organization.
"Our CEO is a former economist who can quickly discern between relevant information and information that is interesting but not as relevant. That trickles down the organization," said Tim Herbert, SVP of research and marketing intelligence at industry association CompTIA. "The investment in [training] and willingness to experiment have been crucial to how we operate."
Tear Down Organizational Barriers
Barriers to information exchange aren't always technological. Historically, departments and operating units have collected and analyzed their own data and presented results as necessary. Mature companies make a point of encouraging information sharing and collaboration to identify and solve cross-functional problems that weren't previously apparent, or to gain a new type of competitive edge.
Allocate Resources Differently
Companies have been claiming to be "customer-centric" for the last couple of decades, and part of the effort has been focused on personalizing experiences and ensuring customer satisfaction throughout the lifetime of the relationship. Today, customer satisfaction isn't simply about customer service. It also involves anticipating customer needs. As a result, workforce allocation is shifting
"We find that workforces skew more toward customer success. There are more product folks than engineering folks in terms of the personnel count because it's more about reacting to data than it is building raw functionality," said Aziz Gilani, a partner at Mercury Fund.
One quick way Mercury Fund assesses its portfolio companies and potential investments is to look at the customer success team, how it categorizes customers, and how it interacts with those customers to determine how well the company identifies, manages, and minimizes the risks of customer churn.
Productize Data
Lots of data, lots of insights, lots of potential opportunity. Sometimes that means supplementing the business strategy with a data product to improve upon how you serve customers, or to create a competitive advantage.
Two years ago, GPS tracking and fleet management software company Spireon was providing its customers with basic reports. It had no real data science capabilities. Since that time, the company has built data science and engineering teams, and launched an industry benchmark data product based on its own data warehouse and the Birst BI platform. Now, its transportation customers can understand fleet utilization, competitively and by location. In addition, the company offers usage-based premiums.
"If you were to ask me if we were data-driven when I joined the company two years ago, I'd say probably not. We had basic reports, but we were not able to do sophisticated industry benchmark analysis and predictive analytics. It has become a significant piece of our competitive advantage," said Ben Tran, senior product manager at Spireon.
