By Sramana Mitra and guest author Shaloo Shalini
SM: What about other functions, say, the mission-critical functions in your business or the areas that are traditionally the first ones to adopt clouds, such as CRM? Salesforce and SuccessFactors are the applications that have built huge companies out of their cloud-based offerings. What is your perspective on them?
SB: We are looking at Salesforce solution in terms of whether there is an equivalent brokerage application that is in the clouds today. For instance, I used to be in the Future Group, and we had SAP deployments. Today, SAP is another company that is moving very aggressively into the cloud. I can tell you that if we had the cloud option when we moved to SAP, we would have taken it in a flash. In hindsight, we wouldn’t have done our own internal SAP implementation in the Future Group. Internal implementations are a horrible thing for enterprise resource planning (ERP) applications. If it works well on the cloud, we will use those solutions based in the cloud. Traditionally, ERPs are very challenging to size, and that is the real problem. You can never get the size right: either you have overcapacity and you are safe, but you spend a lot more, or you have undercapacity and everyone in the business is yelling at you. A cloud-based ERP offering would be great. We have tried other things as well that are cloud based. There is a company called Bitnami that provides ready-made, cloud-ready images for various applications. Currently, Bitnami offerings are free, but there are other companies operating in this area.
Jumpbox is another that provides fully configured ready-made images of applications that are not otherwise cloud ready, and you can deploy them on the Amazon cloud or the GoGrid cloud in a matter of minutes. You can just take the image, connect it to your own instance or they will provide you an instance, and you are up and running. We have tried some of these applications, but we haven’t touched any business-critical applications and don’t plan to move those to the cloud yet. The biggest reason to not do that are regulatory issues. We cannot put Indian stock market data outside India. The second reason is that most of our critical applications work well as of today, so why touch them? What’s our driver here? We need to have a significant driver to bring those to the cloud.
SM: As I listen to you, one of the questions I’m trying to sort out in my head is related to blue-sky opportunities in the cloud from IndiaInfoline’s cloud adoption perspective. We are doing this series as part of our One Million by One Million program. Our objective is to help entrepreneurs come to terms with new opportunities, open opportunities in the cloud from the cloud adopter’s point of view. One of my thoughts is that in the past ten years, at least in Silicon Valley, where I am, we have had 600–700 SaaS companies founded, and many of them are getting good traction. They are addressing various slivers of business workflows all over the place. Most of their solutions are public cloud offerings. What is going on in India? There is a price point at which U.S. vendors offer their cloud-based products or cloud services. Are Indian entrepreneurs taking on some of these different types of business applications and putting them on the cloud, dramatically cutting their cost structure, and offering the same solutions as their U.S. counterparts at a lower price? Where I am going is, you talked about the learning management system (LMS) that Excelsoft provides to IndiaInfoline today. There are other LMSs in the market that you may have evaluated as well. In terms of the strategy of this Indian company you are working with, are they dramatically changing the cost structure, thus making it more attractive for Indian businesses than picking up something from a U.S.-based vendor?
SB: The Excelsoft cloud-based LMS offering is not a dramatic shift into cloud. It would not have cost us much more to have taken that offering as a physically hosted service. The primary difference in cost would have been the hardware that we have to buy upfront. Here, in the case of the cloud-based offering, we are not buying the hardware. We are renting it in small bits and pieces. The biggest advantage for us is that we can switch off the hardware when we don’t require it. So, we are paying for what we use irrespective of the peak load. We don’t pay for the peak capacity infrastructure at all when we are not using that capacity. The tests that use the LMS are not run 24/7. They run in well defined and limited slots. So, we pay only during the slots that it actually runs. Overall, it makes it a bit cheaper for us to buy as a service as opposed to buying the hardware and then using a hosted service because we have a lot of users who take the test. If we were to buy the hardware, we have to size the hardware to address peak usage. Dealing with the hardware price would be a problem for us, in that case.
SB: If we do that, most of time the hardware would be idle and unused. But that is why it’s not like a Salesforce model where I just pay a per-user fee and nothing else.
SM: What is price you pay?
We also use project management software that is integrated into Google; it’s called Manymoon. That is again a per-user license. We would love to use more such cloud-based offerings, but the real challenge for businesses in India in terms of pricing is when you deal with dollar pricing. In terms of dollars, the per-user pricing for India is not as attractive today because in India we tend to put more people on the same job with less automation. There is a much larger turnover in IT, and to deal with those challenges we hire more people. For such IT environments, the challenge in adopting per-user pricing is how to get it low enough so that overall you make a fair amount of money. You can’t charge very high per-user fees for cloud-based solutions in India.