Thursday, January 25, 2007

Does SaaS really stand for 'SOA and Ajax as a Service'?

I know, SaaS stands for Software-asa-Service. But it seems there is a recent and fast moving resurgence (which some call SaaS 2.0) being driven by many forces all coming together. And if we’re accurate, the forces driving SaaS 2.0 are Ajax and SOA. I always thought of SaaS 1.0 as an ASP that represented hosted apps, not really services. And especially not multi-tenant apps. But, this is starting to change in a big way.

Much of the credit should go to Salesforce. Not only did Salesforce stick with its “No Software” value proposition throughout the so-called tough Web-years, the company also opened up its services so third-party applications could too be delivered over the Web. This of course is called SalesForce AppExchange. The corollary to AppExchange in the consumer space is Amazon.com. Some execs at Amazon may have the grand vision of running all Web commerce consumer transactions through their Web services, but I’ve heard them say that they expose (latest count is ten) their Web services because they think developers and companies would find them as useful as Amazon does. Pretty forward-thinking to me.

This is only the beginning. If you follow the Long Tail, Amazon and Salesforce.com should eventually generate far more income from third parties consuming their services than from their own customer-facing software. A while ago I blogged about this and called it the Composite Company. A Composite Company is one that has built its business completely on other companies’ services. The interesting part is that applications are built by consuming services from multiple companies, not just one. When I wrote this, the term “mashup” wasn’t widely known. But, it really does represent an enabling architectural style to let developers and users integrate data from the disparate services. This is the only way it will work and the only way it will scale. Can you imagine if IT had to pre-integrate all services from all companies? Not gonna happen. When made available as governed, secure and reliable (businesses need these), these SOA services – which are truly built for consumption – will create an innovation playground to drive new Web applications that were never imaginable before. But SOA is just middleware, it’s not the whole solution. It’s Ajax that provides the “face” or the last-mile that will enable true Web applications to provide these solutions.

This all will evolve into a new market called the SaaS Consumer market and will be driven by three types of companies:

  • SOA-only Software Providers - A new generation of ISVs that only provide services, but not UIs. Think SugarCRM and Salesforce as a set of SOA-only services.
  • SaaS Market Enablers - Companies that know how to host, manage and secure services for application consumption.
  • Composite Company - A new generation of ISVs that provide applications and composite apps that are completely built by consuming SOA-only Software Providers running on SaaS Market Enablers datacenters.

Get ready for this. Not only will this cause a dramatic shift in how we develop Web-based software, but it will create a whole new market of company opportunities.

Read More...

Thursday, January 18, 2007

It's All About the User, Stupid!

Let me make a less-than-bold statement: the most important part of software is the user. So why am I on this subject, you ask? Don't I sell 'Enterprise Web 2.0' software? That's Ajax, SOA, mashups, etc., right? Sure enough. But I sell software that supposed to empower users.

I've just finished reading a great blog ‘The Peoples Interface Awards' that drives this point home. It's a nice and concise recap of how important the ‘human interface design' is in how users perceive systems (and their IT department).

As a software provider in an industry that makes the web more interactive (and, presumably, more user-friendly), we see the results of bad and good web interfaces every day. Stealing a phrase from James Carville, we often make the statement that 'Its All About the User, Stupid'. In fact, it is such a popular topic in our industry that we've even held webcasts on it.

The issue with this issue is that it is 'soft and fuzzy', without many hard metrics for tangible, bottom-line ROI. Sure, we have examples. We've seen good improvements in customer conversion statistics for e-commerce applications were the ‘Web 1.0' interface was poorly built. But such stats from a vendor like JackBe could be viewed as biased (as we did the work). So it's great to see that this issue is being given some CIO-level attention.

Long live the user.

Read More...