It feels like SaaS has been a story waiting to happen since around 1999. The network is the computer, thin clients with fat pipes, scaling servers, clustering, virtualization... Seems like all the pieces are in place, the cloud is up there. But still there are only a handful of winners that have really figured out the SaaS model, and a sea of also rans that got run over along the way.
Why?
Well, some of them were just plain bad ideas. Irrational exuberance and all that. But I think there is another big factor at play here. When you have these giant, sea change moments in the way that software is designed, built, shipped and supported, not to mention sold, you'd better have the tools and technology to support you along the way, otherwise it's not going to be easy! In fact, it's going to be hard. Really hard.
I could point you to a room full of ex-CEO's who will attest to this fact. Whipping up your latest Web 2.0 mashup and putting it online is usually about 3% of the challenge. What happens when people actually start using it? Here's where the rubber meets the road.
One of the essential elements of success is getting a solid, scalable application online and running smoothly and securely. But there just hasn't been a lot of innovation here.
Being able to quickly identify, respond to and resolve issues in a SaaS application is critical, because if one server has a bad day, it's not one customer that feels pain, it's hundreds or thousands. And that's bad. SaaS acts like a big hairy amplifier on any defect or scalability issue that might be lurking in your app.Technologies like Introscope, Patrol, Vantage, Snort and my company Replay are starting to address the needs, but our customers are still pioneering and forging the landscape as they increasingly feel the pains of this new software paradigm we find ourselves in.
So great job, VMware, Amazon and Linus for getting us to this point where we finally can explore the dream of The Cloud. Now it's up to companies like us to make SaaS applications manageable, cost effective, and safe as we keep things running up there 24/7 with less 'unscheduled maintenance'!
