Sramana Mitra: Switching gears, what are some unsolved problems? What are pain points that you’re hearing from your customer base that are interesting problems and that you and your peers are not solving? These are emerging pain points that need a whole new entrepreneur to go after.
Momchil Michailov: That would be me writing up a checklist for our future competitors. Playing in the infrastructure layer exposes us to a lot of needs customers have that are fairly unique. It’s a little bit difficult to pick a trend and generalize on that. The general notion is customers do a lot larger data sets. For these data sets, they need an enormous amount of performance, specifically around big data and data analytics.
There are very substantial challenges around the mixing of primary data and data that you run analytics on. We see a lot of customers that run enormous data sets. These data sets are very compute and storage intensive. That, on its own, almost overloads their infrastructure. At the same time, they need to run analytics. They need to understand their business. They need to track their inventories and do a lot of manipulation on that data. It’s almost impossible for them to run these types of analytics on top of their primary production data.
The challenge of replicating hundreds of terabytes of data from one data store to another data store and then run analytics on it are both cost and time prohibitive. That is a common issue that we across a lot of our customers. They’ve addressed all of the workload migration pieces. They’ve addressed all the ongoing day to day transactions but the customers that run a lot of big data information systems that need analytics and the ability to both use the exact same date set for ongoing operations as well as data mining are the ones that are suffering the most right now.
Sramana Mitra: You don’t believe that there aren’t people working on that problem?
Momchil Michailov: I’m sure there are people working on it. Fortunately or unfortunately, we don’t know what everybody is doing. That is an underlying common problem that we hear from most of our customers and we are yet to hear anyone use a solution that they’re happy with. I’m sure somebody’s working on it. One of the things that we also see is there’s a lot of these solutions in the web tool architecture. But again, we are also dealing with enterprise customers that are running legacy systems and so it’s not easy to move everything to a new system. It’s not easy for them to move into the cloud. They want to have that on premise. They want these capabilities to work with legacy applications with a lot of data sets associated with them. That is not something we have seen tackled with.
Sramana Mitra: That’s interesting. That’s exactly the kind of answer that I was looking for to that question. Thank you. That’s excellent.
Momchil Michailov: No problem.
Sramana Mitra: That was really high quality. Any other nuggets?
Momchil Michailov: We could probably go through a number of them. When we start looking at that enterprise architecture, the biggest thing that we see from customers is the question of to-cloud or not to-cloud. Enterprise customers are very opposed to it. I was involved in a discussion yesterday and somebody said, “Software, at this point, is free. For $40, you can get a mail client for a year.” They call that free and it’s amusing. If you take a 100-people company, the perpetual licensing that you would normally purchase is no longer available. Now you have this ongoing yearly rent of mail as a service. Ultimately, it’s very far away from free. Customers have hang-ups with the public cloud in general.
One of the other huge issues we see is if customers can get over their fear of the public cloud – security and point of control, service levels and support – there is really nothing out there today that allows customers to span on-premise infrastructure with cloud infrastructure in an efficient and a well performing way. In 80% of the transactions we get involved in, it’s not necessarily a mandatory checkbox that customers have that capability, but customers are desperately looking for it. Today, customers face either to-cloud and use the infrastructure as a service or to build on-premise. If they’re building on-premise on a traditional architecture, they’re facing millions of dollars on upfront cost and they absolutely have to over subscribe because that type of infrastructure doesn’t have elasticity. Or they need to move into the cloud. When they start looking at the cloud and they get all of the uncertain pieces or they consider the legacy applications that they have to move into the cloud, it’s an either-or kind of question. The demand for spanning capability between on-premise infrastructure and public cloud infrastructure is one of the other areas that customers are very eager to find a solution to. We have certainly not seen anything even remotely close to what customers actually need.