Is this the end of IT as we know it?

Halsey Minor, CEO of hosted integration provider grand central Communications, has a powerful message for IT: “In four years, ... basically the whole notion of enterprise application software is going to be dead.” He believes application functionality will instead be available as hosted, pay-per-use services delivered by companies such as Salesforce.com. Putting his money where his mouth is, Minor has recently launched a $50 million venture capital fund with his own money to fuel on-demand startups. For its part, Grand Central will handle data and process integration between enterprises and multiple on-demand services.

Marc Benioff — CEO of Salesforce.com, which originated the “no software” marketing campaign — offers a similar view. “Enterprise software is dying out,” he says. “Look at companies like IBM, which says things should be delivered on demand, and Oracle saying things should be delivered on demand. Even Siebel, who for years and years said it would never happen, is now saying it has to happen.”

Bold talk. And when taken to its logical extreme, it could prove more than a little alarming for enterprise IT. One premise underlying the death-of-software argument upheld by Minor, Benioff, and others is that enterprise IT is drowning in complexity. The business side is tired of big licensing and hardware investments, endless software deployment cycles, random outages, and regulatory-compliance horrors. Rather than keep suffering all that, enterprises will pull the rip cord and opt for the hosted, on-demand model instead — or so the argument goes.

But if companies commit to that strategy wholesale, won’t enterprise IT be subject to massive layoffs? Will those poor souls be sending their resumes to the likes of Minor and Benioff ?

On-demand proponents are quick to mute such overtones. “Those IT folks and those resources are going to be reallocated to the stuff that adds value to the organization,” says Jason Maynard, senior North America software analyst at Merrill Lynch, which last April launched an on-demand index that tracks both pure-plays such as Salesforce.com and software-by-subscription from traditional software houses such as Microsoft.

Or as John Girard, CEO of hosted content management provider Clickability, puts it: “Insource the core and outsource the rest.” But why should IT trust on-demand providers to handle “the rest,” when just three years ago so many failed so spectacularly under another name, ASPs?

The answer is that, although the on-demand model is still evolving, many of the problems that scuttled the first round of ASPs have been or are being solved. Web services have helped make customization and integration easier. Identity management is bridging user provisioning between provider and enterprise. And various technologies — much of them developed by a handful of ASP survivors — are making hosted provider platforms more reliable, scalable, and secure.

Click for larger view.

Such advances lead many to identify on-demand software as the next big thing, although the number of current deployments is tiny compared with the software market as a whole. “In spite of all the hype around it, software as a service is the biggest thing to happen in software in 25 years,” Girard says. “It’s an enormous paradigm shift.”