During Dreamforce 2008 Salesforce.com (SFDC) spent a great deal of time and effort talking about their development platform Force.com and "native" applications deployed on it. This is not a new strategy of course, they have been talking about this topic in ernest for 2+ years. The native ISV talk is new, but not surprising even though most of the current development on Force.com is not native apps, but more feature and function extensions to the CRM apps by customers / SI's. In a lunch discussion at the ISV meeting during Dreamforce I had an interesting discussion around development on PaaS, custom applications, vertical applications and current thinking around customer buying patterns in enterprise software. The discussion centered around the value proposition of PaaS and customer desires for more and better fit "out of the box" functionality targeted at the customer's specific micro-vertical needs and processes. The traditional wisdom story that I defended at the time was that in the 1990's customers accepted more horizontal ERP and then customized that software to fit, but that after the post Y2K recoil customers were demanding more out of the box vertical functionality and were less and less willing to customize the software, little lone build custom software except when absolutely necessary. If the 1990's fit was maybe 50-60% out of the box, I would have guessed that today, with all the pressure from customers the fit out of the box for most verticals was more like 75-80%...so the last 20% or so of fit is still customizations (or configuration to use the term of choice). The premise of the discussion was that in fact the fit for most customers today is really only about 60%, not much improvement in reality and that the real opportunity with the new PaaS offerings is to offer a completely new way to solve this issue (well, you might argue that this is an old way of dealing with the issue just enabled with a business model / paradigm change). The end result of the conversation was that I walked away going hmmm. Now Hmm is actually good for me, since I am supposed to find, track, explain new trends and / or shifts in the industry.
So I've started kicking this around a bit in my head over the last few weeks and while I don't actually have a resolution yet, it is something that bears some more investigation. The first question in my mind is what is the value of PaaS and why is it different from other custom software development tools (if in fact it is different). So what's different (at least what do the vendors claim?):
-more rapid development / deployment
-SaaS like benefits (no infrastructure to maintain, no hardware to buy, easily scalable, etc.)
-lower cost apps once developed (now that would seem to be true based on the Force.com model anyway)
-because of the architecture the old hard to maintain custom software argument is less important (I'm not sure of this one either, but)
-near 100% fit and often accomplished in the same or less time that package software deployment (the theory is that package software that has to be heavily customized can't be deployed as fast as the PaaS based custom software)
Now, like I said, I don't have this one figured out yet, but the argument is interesting, possibly very interesting and if true, could be a very disruptive trend. From an ISV / SW partner standpoint it's a great opportunity to change / build partner businesses. For the customer's perspective, if the value prop is true (or even partially true) this could be a very compelling way to consume software that is a much better fit (and a competitive differentiator). From the PaaS vendor standpoint, well, that's obvious.
Most experts agree that the SW landscape will look different after the recovery from the current economic crisis. I'm convinced that the business and business models are in a heavy change cycle. I think, for example that there is clear evidence that a hybrid business model will prevail (SaaS, on premise and Software as an appliance), simply based on the concept that customers want choice and flexibility. So if PaaS does offer such an advantage for vertical app development how would that effect SW vendors? Well, if the consumption model changed based on this idea then SW vendors would need to product more horizontal apps that support integration to custom PaaS based apps. Customers would build a mixed environment of horizontal (possibly apps like financials, transactional CRM, etc.) apps and custom industry vertical focused apps build on the PaaS of choice...this needs some deeper research...
