Lawson CEO Harry Debes started -- or rather continued -- a brouhaha that's getting some pixels in the blogosphere, and his position is worth commenting on for both its courage (foolhardy) and its excess (hyperbolically so). According to Harry, the SaaS market will collapse in two years time, due to the Saas model's inherent lack of financial merit. Speaking from the perspective of having predicted the demise of SaaS-market leader Salesforce.com a little more than a year ago, and having then had to change the timetable for my prediction (though not the prediction itself), I think Harry's comments deserve a little comment from someone who's been there, done that (sort of).
I think Harry is right that Salesforce.com's success doesn't spell the success of the overall SaaS model, and I would argue that recent evidence of profitability on the part of SFDC isn't enough to keep the competitive and market forces arrayed against SFDC from eating its lunch in the next year or so. While there are definitely SaaS companies that are making money, even good money (as noted by Bob Warfield here), my sense is that pure-play SaaS vendors will "collapse" more because they are ignoring customer choice more than because they are part of a hype curve that's heading south any time soon.
As my friend and fellow Enterprise Irregular Vinnie Mirchandani points out, SaaS will survive because its customer appeal is unstoppable. The pricing, the service model, the pay-as-you-go structure are so rock-solidly pro-customer -- particularly relative to the current on-premise licensing and service models, that SaaS companies can and will thrive for no other reason than this ability to break the current customer-unfriendly model that dominates the on-premise world.
But if Harry had said the SaaS model will collapse because a pure-play SaaS model is, in the long-run, untenable, I would have to agree. These same forces for customer choice will also dictate that hybrid SaaS/on-premise business models will be the ones that truly succeed in the long-run, precisely because customers, being only human after all, want to be able to do two things that, right or wrong, mean some version of on-premise is here to stay.
Customer choice #1 is customization to fit a specific business model: while not all functions need to be customized, the ones that deliver deep competitive advantage to individual customers will not be amenable to the one-size-fits-all on-demand model. Every company has some proprietary process IP that makes it unique, and insofar as those processes are realized in enterprise software, that software will be highly customized, and therefore only available as an on-premise solution. The important trick here is that the same two functions -- CRM, HRMS, ERP, SRM, etc -- can be strategic to one division or line of business and completely generic to another inside the same company, much less across different companies. And those needs change as the business model changes. The successful vendors in the long-run will be able to handle this requirement with the fluidity of a hybrid model -- SaaS only vendors won't be able to meet these customers' needs.
Customer choice #2 is local control over data and process. While we all know the myriad reasons why we don't ever ever ever have to worry about data being stored off-premise (did you catch the sarcasm there?), there are those businesses that insist on having everything locked down inside their four walls. There are even some good reasons for this, but good or bad, if what the customer wants is local, that's the only reason I need to hear. And, because there are lots of companies that feel this way, this guarantees that on-premise won't go away. It also guarantees that the smart on-demand vendors will have a hybrid offering precisely to capture these customers and either convert them when they see the light or keep them from going to the competition.
There are other good reasons why SaaS/on-demand isn't going away -- ever -- such as what I call value-added SaaS, but I think I've made my point. Harry Debes may be right to disparage the relative success of some of the SaaS market leaders, and maybe he's just doing his job by getting more people to pay attention to him and his company than otherwise would be the case, but all in all SaaS will end when customer choice ends, and no sooner. And when customer choice becomes passe in the enterprise software market, I hope I'm long retired. SaaS isn't collapsing, it's only just getting started.