Bold Questions over Big Data: An Interview with Matt Coatney

[interview]
Summary:

In this interview, Matt Coatney discusses the importance of asking bold questions, the big misconceptions behind big data, the best way to start your approach to big data, and his vision for a future where technology and big data make the world a better place.

Cameron Philipp-Edmonds: Can you start us off by telling us a little about yourself and your current role at WilmerHale?

Matt Coatney: First, thanks for the opportunity to talk about something I love. I have always been fascinated with advanced technology, especially how we interact with computers, how they learn, and how teams of humans and machines work together to make sense out of a sea of data. It might seem odd to find someone with such a passion for cutting edge technology working with traditional institutions like law firms, but there is a lot of exciting work going on in the industry. Pricing pressures combined with advances in machine learning have led to transformational changes in areas like document review, billing, and pricing. At WilmerHale, I lead a team of IT management consultants responsible for leading significant change initiatives. Our recent focus has been on what Gartner calls the nexus of forces (big data, mobile, cloud, and social), and we have introduced a number of new capabilities to help our lawyers better manage their practice in these challenging and ever-changing times.

Cameron: You recently wrote an article that was featured on LinkedIn which was titled "Thinking big data? Think Bold Questions Instead." What led you to writing this article?

Matt: I have seen firsthand how organizations can rush to implement the latest technology fad without having a sound understanding of how to use it, and what business objectives and questions they need to address. I wanted to sound a warning alarm for those contemplating significant investments in big data to pause, take a deep breath, and understand first how and where they can use it to drive better business decisions.

Cameron: Your article starts off with somewhat of a polarizing analogy. You take the well-known expression of a hammer looking for a nail and claim that big data fits this analogy in some ways. Why is the analogy appropriate?

Matt: In many industries, most of the data they deal with is actually "small data" or "medium data" at best. In those cases, the questions businesses need to answer are often more straightforward, and they can use traditional tools like reports, dashboards, and trend analysis. Using sophisticated tools and techniques like a Hadoop or neural nets is significant overkill. Perhaps a better analogy might be using a bazooka to kill a mosquito.

Cameron: So, how should someone start their approach to big data?

Matt: As I mention in the article, you should start by understanding what key business questions support your most important objectives. These need to go beyond surface level metrics and get to the heart of what really drives your business. You can think of the Five Whys of Six Sigma or the Lean Startup’s focus on actionable metrics like Cohort Analysis. Once you have those "bold questions," you need to look at your systems to determine what data and tools are needed to answer them. At that point, either a large volume of data or the need for sophisticated data mining can guide you to big data tools. The nice thing is that investment is just in time. You do not spend the money on infrastructure until the need arises.

Cameron: Why is today better for bold questions as compared to the past?

Matt: This is the part that I really get excited about. If you look back fifty to sixty years in computer science, pioneers had many of the same theories and ideas for practical applications that we do today, but two things hampered their ability: lack of sufficient computing power and absence of rich, complex data to analyze. The convergence of exponential improvements in hardware, proliferation of data analysis tools, and increased acquisition and storage of data makes it possible to ask thought-provoking questions and either get an answer or explore the data in near real time, at prices that are affordable to even medium-sized businesses.

Cameron: When asking bold questions with their big data, how quickly can a team expect answers?

Matt: It is really a factor of people and process more than technology. If you have talented individuals that understand the business need, have a can-do attitude, and are empowered with access to the necessary data and other resources, I have seen answers come back in weeks or even days start to finish. What kills momentum more than anything else is well-intentioned but misguided bureaucracy. I have seen the process bog down for months as a group struggled to gain approval for accessing the needed data. The irony is that gaining approval took ten times as long as the time to actually deliver results.

Cameron: Your article concludes with a powerful sentiment and suggestion for people to avoid the hype of big data and focus on bold questions instead. In your opinion, and without diving too deep into another subject, are there other facets of big data that people are either getting wrong or falling short on?

Matt: One misconception is what really constitutes big data. As I mentioned before, what most people think of as big data is not really that big. Transactional data for very large retailers or real-time sensor feeds are the real thing: hundreds of data elements a second, millions a day that just flood traditional systems and hardware. Big data is all about the tools you use for managing such massive data sets and analyzing it with a reasonable turnaround time. The other misconception is around how big data and data analytics relate. People often use these terms interchangeably, but while related they are not the same. Data analytics is all about using machines to help us organize, visualize, and analyze complex data so as to gain insight. This can apply to big data, but it often works just as well on smaller datasets. The reason I think data analytics gets lumped in with big data is that it becomes more valuable the more complex and large the data is.

Cameron: You also have a very relatable vision for the future. You have a vision of a world where technology is a true partner for humankind. Can you tell us more about this vision and how big data fits into it?

Matt: I admit I am an optimist. I believe that advances in AI and human-machine interaction will lead to a significantly better world, helping us solve the most pressing problems of our time. I understand concerns of luminaries like Elon Musk and Stephen Hawking, but there has always been this fear of unknown and new technology, though sometimes with merit. We are on the cusp of a very exciting time, where computers shift from mere tools to partners and colleagues. It is already happening in certain areas. For instance, Watson is now assisting doctors with diagnoses, but IBM found they needed to make Watson respond in a more human and collegial tone to be accepted by doctors. Big data and advances in machine learning and computing power are driving this change, as together they empower more sophisticated systems for interacting with and deriving insights from data.

Cameron: As one final question: You're also an entrepreneur and an inventor; can you tell us a little about your past entrepreneurial endeavors and innovations?

Matt: I have always had a passion for creating new things. Even as a child, I started programming in third grade, just as the personal computer hit the consumer market. That desire to create has led me to start up a number of small businesses, some in this space but others in completely unrelated fields. I also love the outdoors, so we started a luxury log cabin rental business in a nearby state park area. I also helped form and run a number of small law firms. On the invention side I focus on bringing advances in data analytics to new markets like pharmaceutical companies, law firms, and financial services. I especially love the challenge of how to process large or complex data sets efficiently. I continue to invent in this space—my current focus is on how to make sophisticated machine learning more accessible and scalable—and am excited to see what comes next.

 

Matt Coatney is a data analytics executive, entrepreneur, inventor, and author specializing in bringing advanced technology to market for fast growing divisions and early stage companies. He combines a strong business and marketing sense with deep technical understanding to consistently deliver results for his organization and clients. He currently serves as a director for global law firm WilmerHale, where he leads a team that delivers significant change initiatives in big data, social, cloud, and mobile. He is also the founder of Five Spot Research, an applied research company commercializing advances in data analytics.

Matt previously served as a director for the leading legal software provider LexisNexis, where he invented and introduced new product lines in the enterprise search and document analytics space. Prior to that, he was a management consultant for Deloitte Consulting and BearingPoint leading business intelligence and analytics projects. He started his career at early stage company LeadScope, where he invented scalable algorithms to mine chemical structures for pharmaceutical companies. Matt holds a Master degree in Computer Science and a Bachelor degree in Biochemistry from The Ohio State University. He has presented on the topic of data analytics at numerous conferences and has published work in a number of respected academic journals.

Learn more about Matt on LinkedIn: http://www.linkedin.com/in/mattcoatney or contact Matt at [email protected].

About the author

Upcoming Events

Nov 28
Dec 04
Apr 29
Jun 25