Sramana Mitra: If you were an entrepreneur designing a solution, how would design it?
Steve Garrou: I definitely want to pick up on your point on Salesforce and the like. I think you’re right. The groundswell, and the quickest route to market, is definitely focusing on the business unit or a specific problem within a company and solving that problem. That is the quickest way. I wonder if an entrepreneur isn’t wildly successful like Salesforce.com is, it may be difficult to scale. There are opportunities to bridge the gap; if there’s a way to bridge the gap between the IT organization and the line of business with a solution, I would focus there. I think IT organizations are ripe for that type of solution, whether it’s on the software side, on the infrastructure side, or the data side. I think they’re looking for ways to be proactive and meet those needs. But finding an application or service that meets the line of business need and at the same time be able to link it back into the corporate mothership is an ideal place to focus.
SM: Would that be a policy tool, something to manage cloud application purchasing policy?
SG: That’s one area that’s really growing, this area of cloud service brokerage or cloud policy management. We work with a number of those companies that are looking to include their applications into the Savvis cloud or the other public clouds, or even private clouds that exist. I think there’s a lot in that space, and they are finding success. We’re having enterprise clients, large media firms, large financial services firms coming to us and saying, “Hey, can you include this application in your overall solution for us because we like the way that it manages line of business applications or line of business resources. But then, we still get a centralized view.” That cloud service brokerage space is one that’s just now taking off, but we’ll see.
Other things like that, maybe even in the software or the data … the storage spaces, I don’t see a lot going on in the storage space. I know we see things like Dropbox and Box.net, those types of companies, and there’s a lot of backup and recovery for laptops going on, with those applications, with those role-based SaaS applications. But the enterprise is really still struggling there. So, it would be definitely an area of opportunity.
SM: You talked about vendors that are already working on this problem. Would you give us an overview of the vendors you’re working with? What do you see in how they’re approaching the problem?
SG: Sure. There are probably 40 to 60 companies that fit this cloud service brokerage space.
SM: Forty to 60 companies in cloud service brokerage?
SG: Yes, for different levels of capability. So, it’s a fragmented market. We’ve probably talked to 10 to 15 of the top ones, and we’ve worked with just under 10 with various client examples. It’s still an immature market. No one in that space has a complete handle and a complete picture. My sense is there’s a lot of consolidation to come.
SM: Which companies have you worked with? What’s their core value proposition, core differentiation, and so on?
SG: I’m hesitant to name some of the companies, but there are a number of them that will stand out.
SM: I think it would be helpful to get the names of the companies because we are trying to understand a sector with you. That’s why we invite thought leaders to the series, so they can teach the audience about a sector. If you don’t mention the names of the companies, then it’s vague. People don’t know where to go to do additional research.
SG: All right. There are a number of these companies. They focus on different things. There are those that focus on infrastructure management vs. policy and dynamic allocation. There’s a company, Abiquo, in California that focuses more on integrating the management of an infrastructure and go through their policy management, their policy configuration with the rest of the client’s infrastructure.
There are other companies . . . there’s a bunch in this space. I would say the bulk of the 40 to 60 are in this space where they’re looking at dynamically allocating cloud workloads across public clouds as well as private ones. It’s based on a policy, whether it’s an SLA policy or a pricing algorithm, which isn’t quite there yet. It’s all interesting technology, and there are interesting ways to approach the problem. However, what we’re finding is that it comes back to the application. Most applications, as they’re written today, aren’t able to take advantage of the dynamic allocation back and forth between Savvis, Amazon, and Verizon clouds. The applications can’t hold it up, yet. So, if there’s an entrepreneur out there who understands that type of problem and how an application uses infrastructure resources, server storage, computing, those types of things, that would be an area of opportunity. I haven’t seen anybody crack that nut yet.