Jive Leapfrongs SaaS, Lands In Cloud

Are we moving to a post-SaaS world? One social networking software vendor says yes. Let's see if other software vendors--and enterprises--agree.Jive Software announced a cloud-based version of its social enterprise software a couple of weeks ago. When I wrote about it, I thought the company's long-term goal was to push users toward a premises version of the product.

I came to this conclusion for two reasons. First, Jive doesn't have a multitenant SaaS offering, which seems like a major oversight for software companies these days, particularly in the social networking and collaboration market. Instead, Jive's cloud version, called Jive Express, delivers the application and database via a virtual machine that runs on Amazon's EC2 platform. It's not a traditional multitenant offering because the application runs as a single instance inside the virtual machine.

Second, Jive's cloud offering had a 500-user limit (since expanded to 1,000). Given that Jive is chasing enterprise deployments of tens of thousands of users, I assumed the company's goal was to hook departmental users inside an organization via the cloud. If that organization decided to roll out Jive enterprise-wide, it would be forced to choose a behind-the-firewall premises version. And that's what I assumed was Jive's long-term strategy.

Turns out I was wrong.

Instead, Jive has leapfrogged the SaaS model in favor of federated private and public clouds. The goal is to provide software to customers in a private cloud, in which the app and infrastructure resides on the customer premises, as well as on the public cloud. In the long term, Jive envisions customers running multiple instances of the software. "There's be one in the public cloud and one in private and the user doesn't know the difference between them," says Chris Morace, SVP of products at Jive.

In the meantime, the company says the 1,000-user limit for the cloud offering is artificial, so that it can "get its sea legs," with the cloud, says Morace.

"Over the course of this year and next, you'll see limitations between the premises and cloud disappear, and we'll introduce many new features and functions in the cloud first because we can test it there more easily, and then move it to premises," he says.

A POST-SAAS WORLD? So why skip SaaS? According to Jive, customers looking at large-scale Jive deployments weren't that interested. "Multitenant solutions weren't getting true enterprise penetration," says Morace. "There were concerns on data comingling, and on infrastructure lock-in." He says companies didn't like having data out of their control, and if they wanted it back, it would essentially be a flat file data dump.

In theory, with a cloud architecture, the application and data can move from the premises to a public environment with a tolerable level of technical aggravation. That's the case with Jive Express: customers who don't want the app in the cloud can essentially download the entire stack (application, database and data) onto their own hardware and keep running the application. You can't do that with SaaS. (Note that I think Jive's app migration isn't as easy as it claims, but that's an argument for another day.)

It seems to me that Jive positions SaaS as a compromise between unweildy application service providers and an agile cloud. SaaS vendors that wanted to approach customers with a cost-effective service delivered via the Internet had to adopt the multitenancy architecture because the cloud model didn't exist yet.

Jive believes the best approach is to leapfrog the SaaS stage. ""We do ultimately believe that if you fast forward, everyone will be on a cloud, whether private or public," says Morace. "The efficiencies and economics make too much sense not to happen."

So what does that mean for SaaS? "I think the enterprise will adopt around virtualized clouds, and multitenant will have a sweet spot for SMBs," says Morace.

I disagree with Jive on this point. Salesforce.com, Workday and others have demonstrated that enterprises will adopt SaaS, despite concerns over security and privacy. The SaaS approach isn't and won't be limited to the SMB market.

What's more interesting is the question of whether the architectural differences between SaaS and the cloud will affect SaaS vendors in the future if private/public federation catches on.

That's a big "if." Aside from technical issues of moving data between private and public clouds, as well as ensuring appropriate security and service levels, there are platform issues.

Jive's cloud offering runs on Amazon's EC2, which relies on Xen as the hypervisor. But most of Jive's customers run on VMware. The company is doing its own work to make sure the two platforms can play nicely together. "We feel like with our own lifting and coding, that we could make federation work at the end of this year," says Morace.

But I do think Jive may be on to something. Assuming the federation kinks can be worked out (a significant assumption), if I were an enterprise contemplating a new business application, would I prefer a cloud model that lets me move the app and my data on and off premises as needed, or a SaaS model that makes app and data extraction a much greater hassle?