Driving the Data-Driven Transition

Driving the Data-Driven Transition

As companies mature, it’s imperative to build a culture of data that starts at the highest levels of leadership. Buy-in from the top down is often the key to unlocking the headcount and the support you need to start projects that make a measurable impact.

Just ask Balaji Vijayan, Product Manager for Machine Learning at LeanPlum, a mobile marketing platform based in San Francisco. He’s spent more than a decade in the tech industry witnessing the data-driven transition firsthand, and he’s obsessed with helping companies answer the question, “How can we win with data?” We sat down with Vijayan to discuss how companies can kick-start their transitions to becoming truly data-driven organizations.

“As software is eating the world, we start to see this shift now with machine learning,” said Vijayan. “It’s one thing if you’re in Silicon Valley, but you’ve got tons of companies around the world that are also waking up to this paradigm shift. They’re realizing that data is extremely powerful, and there’s a lot of money to be made.”

Data Culture Comes from the Top

Like other change initiatives, becoming an analytics-driven organization is most successful if it starts at the top with the support of the company’s most influential executives.

“Everything will flow from there,” said Vijayan. “It becomes much easier to get budget, and it becomes much easier for you to have a conversation about, ‘Here’s things I’m finding the data, and here’s the benefits I’m going to get out of it.’”

He points to the example of DoorDash, the Silicon Valley food-delivery startup that recently tripled their valuation within five months. Vijayan argued DoorDash stands out because of its founders’ backgrounds in machine learning and because the company focuses on using data to optimize its operations wherever it can—setting itself apart from other Bay-Area competitors who invest more effort in marketing and branding, for example.

Capturing the “long tail” of the customer journey with data can illuminate new ways to drive efficiency, outside of the traditional methods of cutting costs or luring new customers. “You can see there’s two valid strategies to a business, but at the end of the day there’s a reason that DoorDash this year has now come to a $4 billion valuation,” Vijayan said.

The results from a data-driven transition can be dramatic, which Vijayan witnessed firsthand at e-document company DocuSign, also based in San Francisco. A change in leadership at the highest level ushered in new energy around analytics and new opportunities for teams to make an impact.

“If we want to double our revenue are we going to do that just by doubling our sales force? You have to start taking a long, hard look at what things are costing you money and what areas are ripe for investment,” Vijayan said.

Charged with a new imperative to understand the long tail of where sales were coming from, Vijayan’s team dug into the numbers. They spotted an opportunity to target a previously untapped market—insurance brokers. The company then seeded their product with features that would attract brokers and revenue jumped, all by realizing the potential that was already there.

What Starts the Shift?

Assess Your Capabilities. It all starts with having the right infrastructure in place. But depending on your business’s focus and priorities, you might not have the talent in-house to build an analytics infrastructure or even hire data engineers, and that’s okay, said Vijayan. Think hard about the costs and benefits of building vs. buying the solutions you need. Small companies and firms may find it easier to invest in a user-friendly, plug-and-play tool that delivers you meaningful data without a complicated setup.

Recognize Your Company’s Limitations. Positive change can happen when leaders realize that they don’t know what they don’t know. While at Postmates, Vijayan saw this up-close, when the rollout of a new pricing model resulted in an unexpected gap in gross margin that was showing up in P&L statements. A senior leader in a finance role was puzzled but realized it could be a data-related issue and turned to the company’s data team for help. Analysts quickly realized a defect in the code was causing the problem, and they fixed it. By the next month, the gap had disappeared.

That success happened in large part because of the culture, Vijayan said. “We had very good leaders who may not have came from a data or analytics background, but based off of the conversations we’ve had with them and the problems we were trying to solve, they realized that the ways we’ve historically tried to solve this problem weren’t working. Let’s try something new.”

Find Undeniable Efficiencies. Quantifying the cost of wasted time and effort can be a strong motivator for executives who are reluctant to spend on the right data capabilities or infrastructure.

In his current role at LeanPlum, Vijayan convinced leaders to hire a staff-level data engineer when he shared that the process of collecting data to train their model was taking 26 hours. Data scientists would have to occasionally check in while it was ongoing to fix errors, ultimately wasting valuable time they could have spent elsewhere.

The new hire made an immediate impact. “This man has forgotten more about Spark than I’ve ever learned,” joked Vijayan. Because the new team member knew exactly what problems to look for, he got the process down from 26 hours to four minutes, rapidly accelerating the pace at which the company can now build out new machine-learning features.

“It was very evident to [senior leaders] that we had made the right choice in bringing this person aboard,” Vijayan shared. “They knew that we were going to expect to see an increase in our velocity in terms of how quickly we were shipping features, because this person was going to solve problems that we had never solved.”

In the Absence of Data-Driven Leadership . . .

Visionary, data-driven leaders are not the norm at every company. And you might never be able to convince a stubborn executive who is determined to never see the value of data. But in situations where you sense a willingness among leaders to learn, you can start the leadership mindshift by filling in that education gap. Create a proof of concept that fits within the company’s existing strategy and provides true value in the form of saving time or money. This can create trust, which eventually enables you to justify further investment in headcount and infrastructure.

Lastly, you may have enough clout on your own to push through projects that you know will result in a meaningful savings or a big impact on the bottom line. If you can, just go for it, Vijayan said.

“It’s the old, ‘Don’t ask for permission, beg for forgiveness.’ If you actually do it and you find what you’ve done is quite valuable, it’s very unlikely that people at the company are going to challenge you.”

Get a free consultation with a data architect to see how to build a data warehouse in minutes.
Request Demo
Read more in:
Share this post:

Work smarter, better, and faster with monthly tips and how-tos.