Q&A with MapR’s CMO Jack Norris

MapR Technologies is one of the 50 top Big Data startups featured in Startup50.com’s inaugural Big Data 50 report. I included MapR as one of a handful of cMapRLogoompanies whose goal is to simplify Hadoop and transform it from a tool requiring specialized developers to one any business can use.

MapR is backed by $110 million in VC funding and has attracted such customers as Ancestry.com, Experian, Rubicon, and comScore.

I recently spoke with MapR Chief Marketing Officer, Jack Norris.

Jack has over 20 years of enterprise software marketing experience. His experience includes launching and establishing analytic, virtualization, and storage companies and leading marketing and business development for an early-stage cloud storage software provider. Jack has also held senior executive roles with EMC, Rainfinity (now EMC), Brio Technology, SQRIBE, and Bain and Company.

What follows is an edited and condensed version of the conversation Jack and I had in the middle of August. To listen to the full interview, scroll down to the end of the interview.

Jeff Vance: After 20 years of enterprise software marketing experience, you’re at a stage in your career where many people start playing it safe. What prompted you to take the plunge back into the startup world?

jackJack Norris: Actually, I think joining MapR is one of the safest decisions I’ve made. The company has a really strong position in enterprise computing. We’re at the center of perhaps the biggest change to hit enterprise architectures that I’ve seen in my career. It’s an incredible opportunity.

JV: As a journalist, statements like that trigger a degree of skepticism. I hear this sort of thing all the time, no matter the space. Before Big Data, it was cloud that was going to change everything, or it was mobile, or social media, or m-commerce. The list goes on and on.

The truth is I tend to agree with you. Big Data is going to be huge. Arguably, it already is, but from your perspective, what’s the difference between Big Data and all of the other overhyped trends that have come before it?

JN: This is a trend that has been simmering for some time. Back when I was at EMC, a report came out showing that data was growing faster than Moore’s Law. This was around 2004 or 2005, so in the decade since, data growth has accelerated even more. Back in the early 2000s, we had already started seeing that the deployment of storage clusters was increasing rapidly. If you looked at the projections, it was clear that storage clusters were going to get larger and larger and larger and compute farms would need to get larger to deal with that.

What was required to keep up was obviously a change in architecture. Otherwise, it was going to take longer to move data over the network than to perform the analysis necessary to arrive at the results you needed.

We saw that play out with financial services, for example. For doing bond risk calculations, and with a number of other use cases, it took some time for anyone to come up with a good answer to this problem.

Fast forward several years, and when I saw MapR, I recognized right away that they had developed a solution to tackle those intractable data problems I’d witnessed firsthand during my time at EMC and, later, at a storage startup.

The architecture here, driven by our CTO M.C. Srivas, who was at Google and saw Big Data at scale, is just spot on. I think having Google Capital lead our recent $110 million round of funding is the proof point that we did get the architecture right.

JV: From your storage background, you could see data itself becoming a huge bottleneck. Where exactly is that data? How do you get it? How do you import it into various applications quickly enough to act?

JN: You know, it wasn’t just storage. Storage was certainly critical, but there are two other trends coming together to form a perfect storm. First, we have production environments that are completely separate from our analytics environments. That separation is not only infrastructure related, but it’s also separated by a time delay. Often, you have delays of days before you get data to a state where it can be analyzed.

Companies are under increasing pressure to have day-zero responses to address risk, cost, and revenue opportunities on a real-time basis.

The other trend is that as data has grown so has the variety of data sources. The time required to aggregate, transform and take that data and put it into a format that can be analyzed is a huge issue. In the past we were rewarded for sampling data, rewarded for having complex models on small data sets, but, now, in the age of the long tail, having the ability to analyze and look across all the data and easily identify anomalies and exceptional events is how companies are better addressing opportunities and creating competitive advantages.

JV: You mentioned that MapR recently landed $110 million in funding from Google Capital and others. How will the funding change MapR’s roadmap going forward?

JN: I don’t think it will change it. I think it’s a reflection that we have the right strategies in place, and the funding will accelerate what we’re already doing right. We’re already operating in 11 countries today. We have 500 paying customers across industries. We have million-dollar customers in 7 out of the 10 top verticals we target, and we continue to expand our engineering capabilities, our go-to market capabilities, and our infrastructure support.

The funding will help us do more of what we’ve already been doing at an accelerated pace.

JV: Hadoop has become practically synonymous with Big Data, but it has its problems, one of the main ones being that there’s not a ton of talent out there with Hadoop expertise. What does MapR bring to the table for companies wanting to take the plunge with Hadoop but wary of its complexity, its security issues, and its trouble serving up real-time data, or as you put it earlier, achieving day-zero response times?

JN: Those aren’t new limitations. When we evaluated the opportunity early on, M.C. Srivas [MapR CTO] and John Schroeder [MapR CEO], with their understanding of enterprise architectures, decided to heavily invest in enhancing the data platform to address those issues.

What we bring to the table is that we give customers the best of both worlds. They get access to all of the open-source innovations and components of the Hadoop ecosystem combined with a data platform that is suited to enterprise architectures.

One example of that is that the MapR Distribution can be accessed just like typical enterprise storage. Standard file formats are in place, and it looks like a single volume even though it’s a large distributed cluster. You don’t need special training for IT administrators, nor do you need to make changes to existing applications in order to access, use, and leverage the data in the MapR cluster.

JV: It seems to me that you just described one of the main advantages of building on top of Hadoop. It got you most of the way there, and then you could narrow your focus to the pain points that enterprises really care about.

JN: Absolutely. I summarize it as: architecture matters. Having an architecture that supports these enterprise applications is critical. The ability to do random read-write on the storage layer was a big change that we delivered, and it allowed us to incorporate an in-Hadoop database. Then, you have an enterprise database side-by-side with file processing, which really gives developers and application creators a big advantage.

That’s why we’re used in so many real-time environments to optimize revenue, to support ad platforms, and you even see companies like Rubicon who are doing 100-billion ad auctions a day on the MapR platform. These are the proof points showing how MapR can impact a business.

JV: You know as well as I do that Hadoop startups are popping up all over the place. I just finished putting together my Big Data 50 report, which features the top 50 Big Data startups of 2014, and I probably could have focused solely on 50 Hadoop startups if I’d wanted to.

MapR was one of those top startups, but for those who haven’t read the report, what is MapR’s key differentiator? What separates you from the rest of the Hadoop startup pack?

JN: I spoke earlier about the platform. At the lowest level, we provide features that translate into huge performance advantages – three, five, seven times faster – and you can take that speed and use it for faster results, or you can use it to build a more efficient architecture, focusing on cost savings.

We offer complete high-availability, so we can run in lights-out datacenter environments, being used, for instance, in the most sensitive financial services applications. We can do that because we’ve built in high-end features, such as point-in-time consistent snapshots and mirroring, and all the things that IT operations people are looking for. They’re confident that this can be a system of record.

The other thing to consider is the real-time aspects of our platform. Some of the applications that are really moving the needle for companies are the ones impacting business as it happens. With MapR, Hadoop can go beyond being just a reporting platform, or just a data platform for queries and analysis, and can actually be used to make very fast responses to opportunities, whether you’re optimizing ad platforms, whether you’re doing commercial insertions for cable TV, or whether you’re doing fraud detection across 100 million credit card holders.

These are applications that blend production operations with what we’ve traditionally thought of as analytic apps, and, today, time is of the essence for these applications.

JV: Could you give me an example of what one of your customers is doing along those lines?

JN: We have several, but maybe the best example, since it’s been in the news a lot lately, is Beats Music. Beats launched a music service in a very crowded space, competing against Pandora, Spotify, iTunes, and others. They launched Beast Music on our platform, and personalization was their key differentiator – driven by the power of Hadoop.

The scalability of the service was also a key factor. After they launched the service, all of the sudden, demand went through the roof, and they were acquired about six months later for $3 billion dollars.

That’s a great example of how a Hadoop-driven platform can accelerate business value quickly.