TROVE Predictive Data Science isn’t your first technology venture. Are you a serial entrepreneur at heart?
Well, I do love starting companies. The buildup to launching a startup is such an exciting time, but so, too, is growing a company, putting a structure and processes in place around it, and building a winning team. Building a great team and succeeding together is extremely satisfying.
That’s what happened with my first venture, back in the 1990s in the wireless space, which was acquired by General Dynamics. Then again in the 2000s with my second venture, a leader in location-based services.
Is TROVE a different animal for you?
Yes. We’ve built TROVE for the long haul, bringing in Ted Schultz, a veteran and proven leader in the utility industry – a key target market of ours – early on as CEO when it was evident we really had an opportunity to lead this space. We also have an extremely talented CTO, Adam Stotz, who has deep data-science expertise developed working with the Department of Defense. Adam has also built an extraordinary software engineering and data science team, the Science SquadTM.
Having done this a few times now, I see that we are in an important “middle phase” in our evolution, where we have the appropriate structure in place without losing the heart, energy, and excitement of a startup, where each experience – success or failure (though, thankfully the successes outnumber the setbacks) – drives improvement and shapes the company we want to become and continue to work at.
Today, I see TROVE as a company that has proven its value again and again for the audience that matters most – our clients. As such, we have worked hard to develop software – a platform and individual applications, our Solvers – that is scalable, replicable, supportable, and shippable.
How has your sense of the data-science opportunity changed over time?
The predictive data-science opportunity and market have been misunderstood from day one. Earlier this month, I read that GE, maker of the Predix predictive data platform, is considering selling off its Digital assets. While this is big news in our market, the idea isn’t at all surprising to me.
So many technology firms jumped on the “big-data platform” bandwagon, racing to get a data-science platform on the market, but we know from firsthand experience that most clients, even those with huge amounts of data to process, aren’t looking to make a multi-million-dollar platform bet right out of the gate, if ever. They need help understanding the value predictive data science can bring to their business and how they can introduce this technology and practice into their enterprise in a logical, cost-effective way that moves the needle.
Don’t get me wrong. We have a platform, too, a great middleware platform that can scale predictive data analytics across any enterprise. It’s just a mistake to assume clients want to start there. It’s like trying to sell a rocket ship to someone who is just hoping for a quicker, better way to get down the street. At TROVE, we can take our customers to the moon and back, but, 9 out of 10 times, that’s not what they want. We’ve gotten really good at delivering what they want, while also showing them we can help them scale the benefits of great data science as needed.
What else have you learned that is helping TROVE to be the flag-bearer of predictive data science for the enterprise?
Every company selling predictive-analytics software to corporate clients has run into the same roadblock: the sales cycle. It’s just too long.
At TROVE, we’ve learned this isn’t a features and functions sale, it’s a value sale. With “show me the value” as the rallying cry in this space, TROVE has developed an Agile Data Science methodology that has touched a nerve with prospects and clients. When you can work closely with a company, bring in a skilled data science team like we have, and help them start small and stay focused on using data science to turn an opportunity into a benefit or a problem into a solution, you’ve got the makings of not only a satisfied customer, but also a customer for life.
Additionally, we’ve learned that our Platform and Solvers don’t have to be sold “off the shelf”: it is OK for there to be some unique integration and configuration from client to client. We’re beginning to really understand that data-science technology is not perfectly analogous to other packaged enterprise applications.
Finally, we’ve seen the value of building a holistic data science team, our industry-leading Science Squad. Bringing together a diverse, experienced team means we can look at problems and opportunities in a holistic, 360-degrees way. The payoff of this versatility has been huge for us, and we measure it in terms of satisfied clients.
Last Question. What advice can you give the enterprise today about adopting predictive data-science solutions?
First, I’d say understand the difference between “an analyst” and a “data scientist.” The level of sophistication a great data scientist or, better yet, team of data scientists, can bring to an opportunity or problem is an order of magnitude more valuable than what an analyst typically can do.
Second, and this applies to a lot of technology purchases, beware of “shiny objects.” This is a jargon-laden field full of very cool, but unproven, technology from some of the biggest names in technology. Don't jump into the pool feet first. Stay practical. Make sure you partner with providers whose machine-learning and deep-learning solutions support a holistic AI vision for your company.
Last but not least, be very clear about the benefit or value you are looking to gain from predictive data science. And if you don’t know what that is, there’s no shame in asking. Collaboration with your data science vendor up front pays handsomely in our experience.