The Next Chapter in Data Analytics

Data-driven decision making is fundamental for all modern businesses. Companies that haven’t figured out how to use data to their advantage must evolve or risk being outcompeted. It’s been no secret to any VC over the past six to seven years that big data is an interesting and attractive investment area. You may be tempted to think that the area has run its course and all the “unicorn” ideas have been funded—but believe it or not, there are still opportunities to build billion-dollar businesses from an early stage. New company-specific data sources are being created in the Fortune 1000, and companies are collecting 2-3x more transactional information each year. Despite the increasing prevalence and availability of data, Silicon Valley hasn’t fully solved for issues surrounding a company’s ability to quickly glean business insights into which data changes are influencing key KPIs. Until now…

Generating proprietary data alone is not enough for businesses to achieve competitive advantage, the real key to success is in using the data effectively. Too often, we see data as an impediment to speed because the cycles between first asking the right business question, to having a data scientist explore the data and develop a hypothesis (or several hypotheses), to having a data scientist confirm the hypothesis, and finally to having the business function have the right insight to take action, can take days if not weeks.

Why is this the unfortunate reality in even the most tech savvy businesses? Because using data to drive key decisions is difficult and complex—and is getting harder. Coming up with a hypothesis about what’s driving a key metric can be just as challenging as finding an answer, and there are not nearly enough data scientists in the market to aid every line of business within every company. LinkedIn reported that Data Scientist is the most promising job in the U.S. in 2019, with the number of job openings posted on the platform rising 56% since the year prior. Without question, we have a data scientist shortage.

In the early 2000s NEA invested in Tableau Software and witnessed firsthand the enthusiasm for data visualization’s powerful efficiency and eye-popping wonder around data exploration. Perhaps this excitement and urgency to adopt new big data tools distracted the market from taking a step back to ask: Why can’t some algorithm just tell me what is going on with my KPIs so I don’t have to explore to begin with? Sometimes an innovative company idea is just a question away, and nearly a decade and a half later we’ve found the answer. Enter Sisu.

The word Sisu is a Finnish concept representing perseverance, grit, and the willingness to solve hard problems. We are thrilled to announce our Series B investment in Sisu, a company utilizing deep ML techniques to analyze the factors driving key metrics and in return surface actionable results to business users. The cloud-first platform speeds up the cycle from raw data to action, allowing businesses to quickly diagnose customer and business operations metrics and making data powerful again.

We recognize the immense value in a technology’s ability to simplify the complex, and to transform the seemingly intangible into something practicable and usable. We saw the need for Sisu’s solution firsthand, within our portfolio. For example, Conviva, which has billions of sensors around the world for measuring Internet video and some of the smartest data analysts in the world (ever heard of Ion Stoica?)—they instantly saw how Sisu could be game changing for serving their customers by providing faster and more comprehensive diagnoses of what’s driving customer behavior.

Peter Bailis, founder and CEO of Sisu, is what I call a “five-tool” entrepreneur. I adopted this term from growing up playing baseball: the rare five-tool player (full disclosure: I wasn’t one of them!) hits for average and power, has speed, a good arm and plays great defense. Players like this lead teams to victory, and you don’t see them very often. Peter is just as extraordinary: he has high EQ, charisma, unique insight on an opportunity, brilliance and humility—all in one. When we heard he was open to raising additional capital and adding another board member, we jumped at the chance.

At NEA, we love the opportunity to back technical teams, and the potential to partner with Peter is what drove our investment thesis. Peter is well known to the NEA family, having studied at Berkeley with advisors Ion Stoica and Ali Ghodsi (Databricks co-founder and CEO, respectively) before becoming one of the youngest members of the computer science faculty at Stanford, where he partnered with Databricks CTO and Apach Spark creator Matei Zaharia on the DAWN research project. Not only is he off-the-charts brilliant, but he also has an acute sense of business challenges—an ability that has helped Sisu to secure impressive early customers such as Samsung, Upwork and Mixt.

The Sisu origin story is not so different from the aforementioned data-visualization pioneer, Tableau. Pat Hanrahan, Christian Chabot, and Chris Stolte were at Stanford when they developed the core technology which later became the data industry gold standard. We believe Sisu is the next iteration of this data journey—now companies can not only visualize their data, but they can factually understand why it is the way it is. We are thrilled to partner with Peter and the Sisu team as they write the next chapter in data analytics.