Sramana: Let me ask you a question on that, although from a slightly different perspective. I recently had a conversation, last week actually with the CIO of Intel, Diane Bryant.
Frank: Yes, I know Diane.
Sramana: She talked about platform as a service at length. Intel is running its own platform as a service solution in-house. Of course, she doesn’t like to do this. This is not something she is happy about. But her reasoning was that she would not like to be locked into one of the PaaS vendors, whether it is Salesforce.com or Google Apps or Microsoft Azure. So, I would like to get your perspective on what do you see as the platform as service evolution? Where is it going, what are the trends, and how do you see this market shaping up?
Frank: I think you are going to see, ultimately, more and more things move to the cloud, and platform as a service is obviously a natural progression. However, there needs to be a compelling reason to do that. Most technologies, if not all, are successful if they can check three boxes – ‘better, faster, cheaper.’ Many of these technologies that are successful have to do that. You’ve mentioned a bunch.
If you think about it, I am pretty sure you would agree that they are better, cheaper, and faster than the previous incarnation. Having on my music in a music device is actually better, faster, and cheaper than having CDs, which was better, faster, and cheaper than having LPs, which was better, faster, and cheaper than whatever … 45s. So, they are very successful because they check those three boxes. I think you are going to see platform as a service be successful when it can check all three.
When it’s better, faster and the same price, it’s not as interesting. When it’s better, faster, locks you in to their particular platform, it’s not as interesting. Now, if it is dramatically less expensive, we might be willing to make that deal. But I think you are going to have to look at how the environments evolve and what you are getting in return. The other challenge with so much of this technology is a lot of technology has not been architected – this is particularly related to corporate technology – has not been architected or designed for multitenancy. You know what I mean by multitenancy?
Sramana: Yes, absolutely.
Frank: Okay, so I think you are going to see limitation of the vendors as they migrate to offering platform or infrastructure as a service that they haven’t completely figured out, yet, what multitenancy is in the platform in the cloud. I think we are not quite there, yet, on some of that. I think that’s why you are going see things that are more straightforward to do like software as a service.
Some of the collaboration stuff was kind of designed from the ground up, easy enough to migrate to multi-tenants incarnations that are better, faster, and cheaper. I think with some of the other ones, it’s going to be a little bit harder or you are going to see vendors try to put in hooks to make a proprietary. Then it’s kind of like – ‘well, that’s interesting, but I don’t want to be locked in because I might be able to take advantage of these scales curve or the cost reduction curve over time.’
Sramana: I will rattle out a few different things I have seen. This may actually help you also synthesize some of the things we are talking about.
Sramana: For example, one of the types of applications where platform as a service is seeing adoption is what you call maybe long-tail applications inside of the enterprise where people need to create quick applications, maybe even business analysts creating quick applications inside that is not coming in as a full blown software as a service product that you can rapidly bring in the way you talked about the recruitment application that you started with, and subsequent other applications that you have brought in.
These are not yet products available for specific functions, available as software as a service. These are just programs that organizations order custom, that perhaps are long-tail applications. This is one area where I think platform as a service is actually seeing good traction. For example, I am hearing quite a lot of noise that Lotus Notes, for instance, is seeing some difficulty in adapting to the cloud environment. As a result, there is a lot of migration out of Lotus Notes that is happening in that collaboration environment, email, productive apps kind of a collaboration environment. Now, all the applications that have been developed around Lotus Notes in the long tail also need to be then moved out. That’s one of the areas where I think platform as a service is actually seeing some adoption.
Frank: I would come at it a little differently. I think that there is a lot of technology inside companies. I can build analogies in a lot of different directions, but I think what was happening or has happened is this “long-tail stuff,” actually the right answer is to get rid of it.