I recently wrote about how SaaS applications can be delivered as on-premises instances with increasing ease. Now I'm seeing evidence of more overlap and mutual advancement of open source models with SaaS. And a lot of it has to do with the ability to have the application services delivered across a spectrum of hosting options, as well as with more tightly aligned business models.
Just as open source software (OSS) licensed products when deployed on premises can be used as a starting place to offer an upgrade path to commercial products (such as what IBM does), open source can also provide an on-ramp to moving to the SaaS delivery of the OSS applications.
By that I mean an ISV can use open source as part of their development and R&D processes to cut costs and quicken production time. They can make that application itself open source to encourage a community of development for add-on features, bug fixes, and plug-ins. They can make said application code free to download and use, therefore enjoying the efficiencies of viral adoption and distribution.
Then they have some options on how to make a living. They can offer service and support, maintenance, and professional services around said OSS application. They can build out a commercially licensed high-end versions and charge for license fees for on-premises deployment of the high-end application. But now, increasingly, they can also create a SaaS offering of either or both applications (open source and commercial versions) and charge a usage-based subscription fee on a monthly basis. And they can allow the end user organization to then either enjoy that SaaS application as fully hosted, hybrid, or deployed as an on-premises instance behind the user's firewall. Whatever they want.
Increasingly, however, among these various options, the combination of the OSS development and SaaS deployment together in a tango of mutual assured productivity makes the most sense -- for both the ISV and the consuming company or users.
Why? Because the business model of SaaS is such a natural fit to open source development, community, and distribution models. High-volume and low-margin work best when their strengths and weaknesses can be managed across the entire application lifecycle, cradle to grave. What's more, charging as a SaaS subscription can easily replace a maintenance contract, and perhaps even improve the economics for both provider and user. Going from a license model to a subscription model tends to benefit the users, sure, but does short- to medium-term harm to the provider (if they can survive the transition at all).
As a new age ISV, if you're going to make good on your OSS development efficiencies and exploit an open source environment by charging for support, why not host the application and/or infrastructure too? You could charge more per month or year from the account than just maintenance alone. And, at the same time, you can cut TCO costs by more efficiently and centrally managing the application and its infrastructure in production. There are lots of hosts out there to help you. Costs on the hosting will probably continue to shrink.
In fact, OSS plus SaaS offers a compelling way to produce good applications well and cheaply, deploy them efficiently with high availability, and make the customer happy on their own terms -- on-premises, fully SaaS, or some hybrid. Then can pick and choose among these options for an assortment of applications and services, reserving the right to keep internal those applications most likely to be customized and which provide the highest differentiation in the market. The new age SaaS ISV can make a living and under cut those who follow an older path.
Here's the big question for the next five years. How can an application that's built commercially, at higher cost and over longer time as a result, and then be deployed on-premises and licensed at high cost per server or processor, compete with a comparable application that's built from open source code, distributed and refined freely and virally, and then delivered as SaaS on high-efficiency hosted open source stacks and clusters?
More importantly, how can an enterprise that indulges in the commercial lifecycle approach compete in an open market of online services with a competitor that exploits OSS across the lifecycle and passes along its savings in a SaaS-delivered per month low-cost offering? Maybe the provider can even offer the SaaS for free, and monetize nicely around advertising.
I think you know the ultimate answers here. In your gut, you know. Not overnight, but you know.
Two examples of the trends supporting these disruptions will surface at next week's Linux World Expo in San Francisco. Adaptive Planning Inc., a SaaS provider of collaborative performance management solutions, is going to open source its application code. So here's an ISV that competes with Hyperion, Cognos, BusinessObjects, and Microsoft, that's got the SaaS and hybrid deployment model benefits, but is going open source too with its new Adaptive Planning Express version. "it's all about total transparency," says President and CEO William Soward.
In other news, FiveRuns Corp. is making generally available its SaaS or hybrid model systems management service. While they have not fully taken the OSS plunge, they plan to build out a community around the information they can extract on how to keep OSS systems and fully heterogeneous environments running smoothly. Here's a start-up, like Adaptive Planning, that is not just filling a niche, its competing with the likes of BMC, HP, CA, and, yes, Microsoft.
Here are some OSS/SaaS Davids that can actually make a go at Goliaths, and to a viable degree do it faster, cheaper, easier, and as a service. Expect to see other Davids at work next week at LinuxWorld Expo at Moscone.