By Sramana Mitra and guest author Shaloo Shalini
SM: How do you handle pricing for a situation such as the private cloud deployment for the Department of Defense (DoD)? Your primary product is on a per user per month subscription model, right?
LO: Yes, it is use-based pricing, whether it be the number of seats you need or how many Web sessions [you have]. So, a lot our traffic is Web session based for online self-service for, say, chat, social, and community types of products. We tend to sell that on a session basis, which is I think better than on a hit basis because it is how the value is created. Somebody comes, they interact with you; a session is created and closed, so we tend to price on session usage.
SM: How does that change in a private cloud environment where the infrastructure is the client’s infrastructure? You are providing the software layer on top of that, right?
LO: In the case of this particular private cloud for the DoD, we have split that pricing structure. The infrastructures are priced separately from the application usage. The application usage is still the primary use-based, but pricing is really built around the intellectual property, the value that is created from the solution, and not so much on how much hardware and memory you need for a specific deployment.
SM: Would it be fair to say that for a large organization, if the infrastructure cost is brought inside the pricing it is a significant saving for them?
LO: You know, that is an interesting question. Every situation is different. I would say typically not. The reason I say that is because there are real advantages to operating at scale. It is actually one of the unique innovations of cloud computing in general because you have the ability to potentially use a very small amount of a particular feature that is really cost beneficial at scale. The R&D really makes sense in scaling, so if you participate in a cloud, a public cloud, you are taking the advantage of a scale that would be sort of at the back level, [and] all the clients can combine inside of that. The infrastructure costs that you are buying at levels beyond even what a large company would buy for that type of workload.
SM: But we are talking about real innovation that you bring to the table. That is more on the software and intellectual property side, right? So, it seems that you have or you are selling the intellectual property, the differentiated portion that of your offering to clients, especially in this case the Department of Defense, and just the infrastructure layers where they are investing. The question is whether the client in that case is not losing the benefit of your innovations.
LO: No, they are not. But I am fairly confident that from a public cloud perspective, the scale that I operate that infrastructure is significantly larger than what we are doing for just the Department of Defense, so my cost structures are better overall for that computing tier. That is really the only point, in some ways, if you use cloud computing. In general, for an organization to be successful at delivering a cloud solution to a heterogeneous customer base like we do, you have to operate at the highest common denominator. You need to provide the appropriate level of security for the most rigorous of your clients. At the same time, you need to provide the ease of delivery for the least technically capable of your clients, and everybody takes advantage of that highest common denominator delivery.
For support such a mode of operation, our R&D investment up front is significant, whether it be the cost structures of buying homogeneous architecture at scale or the ability to leverage R&D across multiple clients. There are unique innovations in delivering computing in this way. That is one of the reasons we have adopted internally. I think it gets passed on, and I, like everyone else in the consumer [ . . .] will be a consumer of private cloud type things. But there is something uniquely advantageous about public cloud computing, because of those types of innovations coming to the table.