As much as I like open source, there is something to be said for not having to install or maintain applications. But that doesn't mean it's smart or realistic to move all the applications in your enterprise to on-demand delivery.
As Gary Rivlin writes in today's NY Times "few software companies make the move to platform status" let alone the ones that have no footprint on the desktop. This is something I noted back in September after the Dreamforce event which reinforced my belief that enterprise software isn't going anywhere--at least not infrastructure software.
In the case of infrastructure (like networking and integration), and desktop environments (like Windows) it's hard to get excited about doing everything through a browser. Not that it?s a bad idea, but as I learned onit's not entirely feasible at this point to do everything via the internet. Rivlin writes:
And yet for Benioff, the company's chief executive, that is not enough. He wants to turn Salesforce into a platform like Microsoft's Windows operating system, a product so popular that it is the foundation for a veritable ecosystem of software developers.
If you at the on-demand subscription offerings from Microsoft and SAP, both mediocre by comparison to Salesforce.com or SugarCRM (also available open source) the main advantage they have is that the existing user base is tied to a set of desktop applications which reinforce the desktop computing paradigm.
"If you know how to use any of Microsoft's desktop tools, you know how to use Microsoft's CRM product," said Bruce Richardson, the vice president for research at AMR Research, a technology consulting firm. Microsoft is a minor player in the CRM market, but its Office software suite is installed on hundreds of millions of computers
The real question is if any application that doesn't reside within an enterprise can develop a true ecosystem. The rub that I found was the difference between a community vs. an ecosystem as I posted on Infoworld.
With the ecosystem approach there are two possibilities: you make enough money to support the business or the mothership acquires you. The issue for the companies at Dreamforce is that Salesforce doesn't acquire many companies, and if you look at the way these SaaS applications are built they might use the same underlying technologies but don't share a common architecture (schemas, multi-tenancy for example) that can be merged into a bigger app like Salesforce.
Contrast this with the community approach where a company can sustain the business, or be acquired by a much broader range of companies. Witness how Citrix acquired XenSource and Yahoo acquired Zimbra, proving that open source and the community approach have far more options for success.
Cont One other key aspect of community is that open source has both "developers" and "users" where SaaS tends to only have "users" which we should probably term "consumers" since there haven't been many well known contributions from SaaS companies back to OSS projects.
So, what are some of the reasons why Windows has an ecosystem (and a community) and why is it attractive to ISVs (independent software vendors) to develop applications for Windows?
Clear ways to monetize
It's much easier to sell packaged or downloadable software than it is to explain that your product only works once you go through a process through another vendor. With Salesforce, 3rd party applications don't feel tangible. They should almost start shipping CDs.
Clear ways to market
With Windows there is enough history to know how to market your products to the user base. In the SaaS world there are few repeatable patterns beyond SF.com. That, and the fact that you are at the whim of SF.com in terms of visibility to their customers.
Clarity from Microsoft on how the relationships work
You pretty much know that Microsoft is only going to let you get so far before you become a target of some kind. Salesforce hasn't been great with its messaging around AppExchange, Apex or anything partner related. But, they have also been making it up on the fly.