These are the early days for big data, and users of it can be overwhelmed by the challenges of analyzing, storing and securing its oceans of information. That’s where Hadoop, an open-source platform for managing big data, comes in, as well as companies such as Cloudera, which distributes Hadoop and then helps companies profit from it.
The Wall Street Journal’s Rolfe Winkler sat down with Cloudera Chief Executive Tom Reilly to talk about his growing company, Hadoop and big data. Edited excerpts follow:
MR. WINKLER: What do you do with all this data that people are collecting?
MR. REILLY: Hadoop is geared toward the new data that’s coming from the connected world. The connected car, the connected home, the connected patient. The data coming back, mobile, social, does not fit into the traditional tools. There’s a volume and a variety and a scale that just doesn’t fit. Hadoop is designed for this modern data.
What’s important about this modern data? The connected world is either a threat or opportunity for every industry. If you’re not rethinking your business around the connected world, then it’s going to be a threat. Because there are going to be challenges.
MR. WINKLER: Give us an example of how a business uses Hadoop to manage and analyze data in a way that’s making them money.
MR. REILLY: There are three classes of use that we see applied to every industry. The first is customer 360. Can we know more about our customers, know more about their demographics, how they use our product or service, so that we can either sell them more or keep them as a client? The telecom industry is a very, very big user of this. They are trying to identify which of us are disgruntled customers and likely to leave. If a large telecom provider can reduce customer churn by 1%, it’s like $500 million to their bottom line.
[To return to the] three use cases, it’s customer 360, [No. 2 is] optimizing their network so they are now understanding, “Where are most of our users, where do we need more bandwidth, who’s downloading more videos, where are we seeing poor service?” The third use case is taking data about us [customers], where we’re at, and selling it to merchants or financial institutions [who use it] for location-based offerings.
MR. WINKLER: Adoption is pretty low for Hadoop at this stage. Is there too much hype?
MR. REILLY: There is way too much hype. I came to Cloudera about two and a half years ago, and I’m like, “Wow, big data, Hadoop, boy.” I got this opportunity to be CEO. I did very little due diligence. I’m like, “I’m in.” And I got in, and I’m going, “Well, it’s not as big as I thought it was,” because of the hype.
But this market is growing very, very fast. And that growth is because of open source and people’s willingness to download and start understanding the technology. With our partner Intel,INTC -0.52 % we entered China a year and a half ago. We [Cloudera representatives] had never been in China. And what we found was our biggest customers were in China. The technology went there before the people. And so that’s why this is growing so fast is because the ability to experiment and learn with open source allows the skills to grow much faster.
MR. WINKLER: Hortonworks, HDP 7.;