Sramana Mitra: Maybe the one data point that would clarify the segmentation is what is the revenue level range of the 10,000 clients that SunGard services?
Indu Kodukula: It is starting at about $2 billion and going down to several hundred million.
SM: It is not SME; it is more the mid-size enterprise?
IK: That is right. It is very much heterogeneous. I think we are going to see both. We are going to continue to serve this client base and, in particular, they are going to continue to buy infrastructure from us. For the next new services, they are going to buy services from us that we might provide on our infrastructure or third parties. What we are also starting to see isthe effect that you mentioned, which is we are going to enable a whole new set of service providers, the application providers, if you will, that are in turn going to go sell those services to a number of clients that we don’t serve today. This is the model that you mentioned, 500 application providers that in turn reach out to hundreds of thousands of new customers. They are absolutely starting to see that effect as well.
SM: How much of the 10,000 customer portfolio falls in that category of application providers servicing tens of thousands of customers?
IK: It is a small number, but it is a significant portion of the overall volume in terms of revenue. So, for example, one of our large what we call ITOs –IT outsourcers – relationships is with ACS (a Xerox company). I think ACS is an example where they don’t serve hundreds of thousands of clients, but we have ACS as a relationship where they take advantage of our services and in turn provide them for a very large number of their own customers. And should we start offering services on third-party infrastructure providers like Amazon, I think we will start to see a whole new dynamic in this space.
SM: Let me understand this carefully. The ACS, Amazon’s ACS business runs on you. Is that what you are saying?
IK: The recovery business that ACS provides, the recovery work that ACS does for all its clients is delivered from our data centers and on our floor, that is right.
SM: If you were to expand your relationship with Amazon and in the model that we just discussed, what otherservices would you be providing?
IK: Again, let us take the case of an application that needs to be recovered, or a cure of an application that we need to ensure that the down time is no more than 15 minutes. You could certainly imagine that an application is set up to replicate its state over to a third-party infrastructure like Amazon’s, where the events that need to take place when the primary side goes down, that entire recovery, the orchestration happens because of intellectual property that we provide. But throughout that whole process, all the infrastructure continues to be Amazon’s. So, whether it is the compute, whether it is EBS, whether it’s S3, all the infrastructure continues to be provided by Amazon. The application, again, restarts on Amazon, but the entire recovery process, the setup of the replication on an ongoing basis, the migration of the application, or the failback of the application from Amazon back to the data center, all of those are services you could certainly see as being provided on Amazon infrastructure by someone like us.
IK: I think that will open up, if we were to do that, that would open up the price point that would be much lower than what we offer today and will be much more utility or even mid market oriented, given the economics and the scale of Amazon, if they were to try. It would be a new addition to our service portfolio.
SM: I am going to ask one more question on this, and then we are going to switch to a question about entrepreneurial opportunity. If you assume the model is going to be that a large number of cloud application vendors are going to sit on services like yours to cater to tens of thousands of customers, then you need to seed that market, and you need to proactively recruit companies that are building cloud applications. Some of your competitors are focusing on recruiting more startups in that space. Do you have a strategy like that?
IK: Absolutely. I think I’ll break that down into both the technology event side and the commercial side. What we are finding – and we are still relatively early in this – is that by providing assistance from a technology and migration point of view, we are able to go advise some of these next generation application providers on how to build and architect their services so that they can be paid at Internet scale, and how that can result in the level of availability they need for their applications. That is where we are finding most of the help is needed by the next generation application service providers. I think that is an area we are looking at to make significant investments in. What we are finding is that once that is done, the commercials are typically fairly straightforward. If the initial discussion on how we help our clients build their applications and architect them so that they can in turn service a very large number of clients, that is what we’re seeing the most demand for right now.