This BeyeNETWORK spotlight features Ron Powell's interview with Michael Howard, Vice President of Marketing at EMC Greenplum. Ron and Michael discuss how the strength of EMC has propelled Greenplum big data analytics forward for use by the largest enterprises in the world
Michael, it has been more than two years since Greenplum was acquired by EMC. Can you talk about how the power and strength of EMC has affected Greenplum's strategic direction?
Michael Howard: Two years ago, Greenplum was a startup. It had been building MPP systems for a while and getting a good cadre of customers, but clearly did not have the scale to meet the expectations of the largest enterprises, the Global 3000 in the world. Now, as part of EMC, Greenplum has seen a meteoric rise in the number of customers, in the scale that we get in terms of being able to meet the demands of quality, and in the ability to meet with our customers across all the different theaters and regions whether it be APJ, EMEA, or the Americas. But what is really super important today, especially with the big data aspect, is you cannot just do big data when you are a small startup. The prerequisites on the technology side, the consulting side, and the data science side all have to be put together. That is what EMC really affords us the appropriate scale, and innovation that is possible under the EMC rubric.
As I recall, Greenplum had a very good presence in Asia but not necessarily North America. Will EMC be able to take Greenplum worldwide and expand their industry base?
Michael Howard: Yes, and especially when it comes to telecommunications, healthcare, and financial services. Some of the opportunities and things that we are doing for some of the largest companies in the world are not only fascinating because of big data, but also would never have happened under a small startup.
Could you tell us how you define big data and why you feel that EMC Greenplum is uniquely positioned to handle big data?
Michael Howard: Well, as you know, big data means so little to so many. I think we can start with that quip. Traditional definitions start with the fact that big data breaks traditional IT architectures. In that sense, there are new opportunities for displacing traditional database and storage architectures to be able to handle the volume of information that is emanating from so many different sources. That is one basic definition. Now there are certain analysts who talk about the three Vs: volume, variety, and velocity, and some add a fourth C complexity. For me, there is a new definition, and it is simply this: big data means new data. By new data, we mean new insights. By new data, we mean new opportunities. With new data, organizations may finally be able to get to new levels of top-line performance.
Do you feel Hadoop is being overhyped?
Michael Howard: Well, Hadoop is being overhyped in the sense of what it can and cannot do. At the same time, it is a symbolic representation, whether we like it or not, of big data. It is the name that is ascribed to big data in terms of a technology or computational fabric. Now, what people are going to find is that it cannot do certain things. I will give you one example: it is more batch-oriented than it is real time. But I think what is going to happen and what you will see through the partnership between EMC and Greenplum is that there will be new ways to use Hadoop and make it more fitting for the enterprise.
What challenges are today's enterprises facing in their analytics efforts as a result of big data?
Well, one is organizational alignment. You have to approach big data differently
than data warehousing and traditional business intelligence. Ron, when we were doing our thing ten years ago, we had to have a very strict return on investment. We had to do all of our interviews with everyone who had a question of the system before we would even start implementing the system. We would have to make sure all the data was clean, modeled, normalized, denormalized and, finally, in compliance with a variety of IT processes that extended that project to a point that it would meet the current business needs.
In the big data world, its a much different challenge. It is sort of opening up your mind to the notion that maybe you shouldn't have a conclusion before you start taking in all this information. Maybe you shouldn't have a supposition that you're prescribing on the information. Maybe you need to bring in data scientists and figure that out. Maybe the initiative really should be aligned with the CXO rather than IT database administrators.
Can you tell us about the Greenplum Unified Analytics Platform and how that helps your customers with their analytics challenges?
Michael Howard: We were the first to come out and say you need a unified analytics platform to solve the challenges of big data and analytics in general. One challenge is to be able to accommodate different types of data, structured and unstructured. Therefore, you need a certain type of database technology, and you also need to support Hadoop. We integrated those two technologies together so that's unification at one level. But there was another capability that we brought to bear that I think is just fascinating and something that we recently announced at Strata. That is our Chorus platform, which is a data science platform for collaboration amongst data scientists because data science, as we say, is a team sport. Analytics is becoming a team sport. Just to be able to find those nuggets of gold in those mountains of information takes multiple insights from multiple people so collaboration and how it is unified with technology and people is very important here.
With so many different choices for big data analytics appliances, what sets the Greenplum Data Computing Appliance apart from the others?
Well, I think what sets it apart is what I just talked about in terms of a unified analytics platform: to be able to roll in, very simply, our Data Computing Appliance (DCA)
, which houses and supports this unified analytics platform with Chorus, with Hadoop, with our MPP database, and with all the necessary openness to be able to reach out to a partner ecosystem. Whether it is our appliance or our software, we're very open. We're not a monolithic kind of company that some of the others are growing into. I think that's very, very important along with the simplicity, the linearity in terms of scalability, and the purpose-builtness of it. I could go on and on about many different features, but I think those are the key ones.
Today large enterprises are all struggling with what to do with Hadoop. Most of them have it in a pilot phase right now, and they have some data scientists looking at it. What does it take to deploy the Greenplum appliance? You mentioned Chorus. Is it easy to learn?
Michael Howard: If we take Chorus, for example, what we did is we leveraged one of our companies that we recently acquired called Pivotal Labs. They are experts in building big data applications. They were used by Google, LinkedIn, Groupon, Facebook and Microsoft. Pivotal Labs was used to optimally build a user interface that masks the complexity of technology beneath it. We took that user interface intelligence, whether for an iPhone, an iPad, tablet, laptop or whatever, and made that the face of our collaboration platform. So if you are acclimated to using LinkedIn, you can use this. It has the same simplicity, ease of use and workflow that true consumer applications have. As a result, it's like social meets big data, encapsulated in our platform for data science. That's very, very unique.
You mentioned tablets, phone and laptops. Mobile is a major focus with our audience. How do you handle mobile?
Michael Howard: The server clearly can support any device. But again, as I said before, we went one step further than that in assimilating Pivotal Labs' expertise in building applications to allow them to be used by any user across any type of device. If you think of big data, there are three components. There is the application side to make it easy to use and consumable by everyone. Big data should not be relegated to the high priests of IT. It's designed for business people to make decisions. Second, you need data science to ferret out the insights, and you need the right computational platform, the Data Computing Appliance, to put it all together.
Michael, with your wealth of experience in high-tech, what you think will be the biggest changes for big data and analytics in 2013?
Michael Howard: Well, on a personal note, what I love about big data is how it connects the dots with our daily life. Again, going back to our early days Ron, if we talked to our children about data warehousing, and business intelligence, and relational schemas, and DQM and all that other mumbo-jumbo, our families certainly would not understand that.
I've been part of something called The Human Face of Big Data
, which is a very important initiative for EMC using Rick Smolan, a famous photographer, to connect a face with big data. There are so many wonderful stories of how big data really affects people and changes lives. I think that next year we're going to see compelling stories about big data and for the first time connect this large enterprise type of technology with our daily lives it will be the new normal, so to speak. Clearly the presidential race between Mitt Romney and Obama is going to have the CMOs of the world really take a look at how they do target marketing. If there ever was an incredible use case of target marketing and micro marketing using big data to win an election of this magnitude, it was the most recent presidential election. Certainly, those kinds of things are going to seep through into every CMO's mind across the Global 3000 in 2013 and beyond.
Excellent Michael. Thank you for providing our readers with an overview of how EMC Greenplum is addressing the challenges of big data.
Recent articles by Ron Powell