By guest author Shaloo Shalini
SM: Very interesting! What you are saying here is that from the vendor’s point of view, the sales cycle has become a lot simpler because you now don’t have to spend as much money for the pilot or proof of concept, whereas earlier you had to go in search of resources in the client premise to enable them to try your solution.
Now, clients can try your product on a cloud-based infrastructure, which is more cost effective, scalable, and so forth, and see if they like your solution and the technology, right?
NS: Yes, exactly. At GigaSpaces, we have gone through the same process ourselves. When we introduced the cloud at GigaSpaces, it was initiated by salespeople and we didn’t want to switch our sales model. What we really did here was create the ability to try out the product on a real data center which is in the cloud. This allows you much broader spectrum of experiencing the real product before installing it within the enterprise, getting approval, or going through a large sales cycle. All of that before was very expensive – just to be able to try out the product.
The question of business model remains, but it is more of a separate question now. This separation makes it a much easier question to deal with after your customers have experienced the product on the cloud without having to invest time and money for resource deployment during the pilot. Once people go through that experience, they are more aware of the product, and our ability to deal with that question becomes much easier with this phased approach.
SM: Well, what I hear you say is through this approach, vendors have already established the value proposition and technical acceptability of the product that goes toward the technical decision making. Once the value proposition is established, what remains is how the structure of the deal or the contract is worked out. This is interesting; you are kind of breaking down the business model into a new sales cycle.
NS: Exactly, that is the case. We can get the value of the cloud without necessarily changing the business model upfront. We work on the business model and deal with that separately from establishing the value proposition of the solution itself.
SM: What are trends you are seeing in the cloud with respect to integration? There are all of these different cloud applications and services – SaaS-, PaaS-, IaaS-based – with new vendors and new models that need to coexist with on-premise solutions.
NS: I mentioned at the beginning of the interview that the simplest cases and simplest forms of solutions are the ones that show up first on the cloud, because they don’t have to answer the question of integration yet. For the more complex scenario where you already have some solutions deployed and where there are large services that you need to deal with, integration becomes a complex issue.
In one of our customer’s cases, they used hybrid approach to address the issue of integration. They said, We can’t take all our back-end systems into the cloud. But it still made sense for them to launch new applications in the cloud. What they did was to take the Web front-end, which is the of core of their app from a user perspective, and deploy it on the cloud with a secure connection to the back-end systems. This way, all of their users’ experience components run in a cloud, such as the ability to select products and browse options. Once users commit to a transaction, that goes via a secure channel at the back-end and they go through their regular processes. By their use of a hybrid process, this customer could deploy application in a very fast manner in terms of business agility and flexibility, but they did not have to transfer the entire organization to the cloud to make it work for them.
SM: You talked about analytics earlier, so if I am a Fortune 500 CIO who bought Salesforce for CRM and maybe analytics from LucidEra, Birst, or myDIALS, how complicated is the process of integrating these two cloud applications, and what are you hearing from the vendors on that front?
NS: Well, vendors believe that kind of integration is not easy and relatively complex. Most of the analytics will give you fairly low-level service to manage a large data set and the process of loading that data from your CRM or whatever into their analytics environment. Then you process or crunch it yourself. That being said, this area is evolving very fast because of large adoption and mindshare that technologies such as Hadoop are getting and the number of vendors using Hadoop. There was an interesting article in GigaOm recently about the commercial adoption of Hadoop in the analytics domain. We are seeing an increasing number of analytics providers that add or deploy it on my local data center or have integration with the salesforce, addressing such requirements [e.g., Pentaho], because there is demand for it. If you are asking about right now, well, that is a relatively complex process. Are people really doing that? Yes, they do it themselves in some cases, but the majority will not owing to the complexity. In parallel, what is happening in terms of integration, especially for CRM companies such as Salesforce, is that the barrier to entry for such integration is getting lower. What is happening here is that the analytics and services that Amazon provides with Elastic Map Reduce (EMR) create an ecosystem of vendors such as Salesforce to provide that kind of integration themselves and give the ability to strap on analytics that more integrated into the offering. That capability is not there right now, but I think it will happen soon.
[Note to readers: More information on vendors that are harnessing commercial-grade Hadoop for petabyte analytics is available here.]