Salesforce.com wants to be the next Internet. No, that's not the way the company describes its plans. But a year ago, Salesforce rolled out AppExchange, a website for on-demand applications built by other vendors but running on top of Salesforce's own software. This month, the company launched a preview version of Apex Code, a Java-like language for building those apps. Salesforce executives say they want Salesforce to be a platform, not just a software-as-a-service product.
How serious are they about this? Serious enough to cannibalise Salesforce's own hosted applications business. If somebody else builds a better Salesforce on top of the Salesforce platform, that'll be fine -- or so the company says.
That's a little tough to believe.
IBM wasn't willing to do that when its mainframes ruled the IT landscape. Microsoft makes sure its own Windows applications always have an advantage over those of other vendors.
There's lots of money in applications. That's how Salesforce has built everything it has. Why would Salesforce risk a profitable and growing business just to turn itself into a platform and let others make money with it?
It makes no sense -- at least not without the Internet. See, in the mainframe world, you leased everything -- from the infrastructure on up to the applications (which you leased or built). So IBM wanted to lease you as much of that stack as possible.
In the PC world, you buy everything -- hardware and software, from the infrastructure on up through the apps. So Microsoft wants to sell you as much of that stack as possible.
But in the Internet world, you don't buy or lease the infrastructure. Somebody else owns and runs that. And somebody else owns and runs most of the software in the middle, whether that means Salesforce, eBay or a Web-hosting Internet service provider.
You can buy a piece of that infrastructure and run your own Web servers and online applications. But you can get by with owning just the tiniest sliver, only the part that's unique to your business, and leasing the rest -- most of it indirectly.
And the money? You pay at the top, and it trickles down to everyone who owns a piece of the Internet stack. That's Salesforce's dream. Salesforce figures if it can build a broad enough platform, with enough on-demand software providers stacked up on it, then lots of that money will trickle down to it from end-user customers that Salesforce never sees or has to fight for. And Salesforce will collect revenue every time those indirect customers use someone else's application built on its platform. So in theory, Salesforce doesn't need the IBM or Microsoft own-the-stack model.
In practice? Well, revenue is revenue. Ultimately, Salesforce isn't likely to walk away from that lucrative CRM money. But Salesforce's platform dream still matters. Why? Because of all the interesting on-demand application ideas that would never see the light of day if they couldn't be piggybacked on Salesforce's platform -- ideas Salesforce would never come up with but might help along.
Most of them may be junk or useless. Some of them may show some value. One or two might even fundamentally change the way you do IT.
That's makes Salesforce a lot more valuable to corporate IT as a platform -- and as an idea incubator -- than as an application.