‘SOA’ Initiatives Tend to Fizzel After Year or Two

‘SOA’ dead as of January 1st, analyst says | Service-Oriented Architecture
| ZDNet.com

"One of the challenges with SOA is that a successful implementation is a multi-year process, since not only do systems and interfaces need to be changed, but development methodologies and incentives need transformation, and the business itself needs to drive the process.� Definitely not overnight stuff. As I understand from what I’ve heard about Burton Group’s own investigations, SOA initiatives show a lot of potential when they are first launched, but tend to fizzle after a year or two."
This seems to be more of a commentary on the difficulty most businesses have in maintaining focus and momentum on strategic IT work...

Do Not Let the Perfect Be the Enemy of the Good!

This post's title is cribbed from Celent's Latest "Report on SOA in Insurance 2007". The report is primarily the result of surveys conducted by Celent with a number of insurers and includes mid size to large carriers (premiums > $1 billion). The report is mildly interesting for someone in the midst of deploying SOA techniques/technologies in the insurance industry (namely me;) but clearly reiterates three key points:

  • The promise of a fully realized SOA enabled infrastructure is alluring and there is an incredible amount of hype around the potential.
  • Building out a fully realized SOA enabled infrastructure is hard and will take time
  • Even small steps toward SOA can pay dividends so get started now!

I Think the quote below best summarizes this:

Celent - Web Services and SOA in Insurance 2007
While the model may not be very sexy, Celent believes that for the next two years Web services/SOA in insurance will remain essentially a "Plumbing" issue - a set of technologies and practices that make it more efficient to share data and transactional capabilities between systems, both internal and external, in a reusable way that allows the value of systems investments to be leveraged repeatedly in subsequent initiatives. While this may not get the CEO's attention, it is absolutely key to the CIO's mission to do more with less.

So get started... Each company is unique in its own way and requires it's on approach but don't get caught up in analysis paralysis. Do not wait for the perfect project or tool. Starting allows you to "do" and doing allows you to learn from the experience. Learning allows you to adapt and "do" again providing opportunity to continually improve as you go. "Do not let the perfect be the enemy of the good!"

Mega Services: Getting There?

It is always interesting to hear Steve Ballmer spin Microsoft's latest marketing, product, technical, whatever announcement. As if on cue eWeek.com published an article titled

"Microsoft CEO Touts Company 'Bet' on Web Services"

Quoting Steve Ballmer characterizing the move as a "big bold bet". I just had to laugh. How could anyone view this as a "bet" in the current technology environment. About the only part of this decision that could be considered a risk is weather to use the W3C "formal" Web Services specifications or a more lean REST based architecture. Granted the formal Web Services specifications are starting to get that hairball consistency, but there are still valid reasons to go either way.

The article goes on to state...
Microsoft believes Web services will work in tandem with PC-installed software, a vision that differs from that of "software as a service" advocates, such as Salesforce.com and Google, who expect services delivered over the Web to replace traditional software... "We believe this shift is the most important technological transformation during the next decade," Ballmer said.

Aahhh.... now this is more like it. I began to have a flashback. [the screen goes fuzzy and "Dreamweaver" begins to play in the background...] In my days at Viant Corp. during the Web 1.0 heyday (don't remember if it was 98 or 99), I had the privilege of being invited to Seattle for a special Microsoft conference. Before I arrived, I didn't have any other information than that it was a selected audience of internet savvy technology partners.

Microsoft began by discussing what they felt was a challenge for them. They were encountering both resistance and lack of interest to using their products any deeper in Corporate America than the desktop. The biggest issue seemed to be the integration challenge. Unix seemed to be the OS lingua franca for integration projects and Sun was selling servers by the boat load.

Microsoft proposed that they were considering XML messaging as the lingua franca for their platform and were seriously looking into embedding its use throughout all its products. From applications like Excel, Outlook right down to the core operating system, all would expose access to features through XML messaging. I was not a Microsoft fan at the time, but the idea was exciting, Microsoft even had a cool code name for it... "Mega Services!".

Alas, although Microsoft bore SOAP and .NET, the vision never seemed to fully pan out in the way described in that conference. This latest announcement is probably the closest they've come and is a step in the right direction, but I still hope for a day when the vision is fulfilled;)

Business Enabled Integration...

A few years ago I worked at an incubator called 12 Entrepreneuring. It was during the heady but waning days of the internet boom. 12 was in the midst of incubating Grand Central Networks, an integration network company. I was asked to help temporarily with the conceptual design and architecture on the project, which was based in the San Francisco office, until the New York office (my eventual homebase) was up and running.

My background was in systems integration at the enterprise level. I felt that the value of the network would be to provide a rich set of pre-integrated business partners and industry aware components. My point was to align with the business and not be purely technology focused. Founder Halsey Minor pushed to keep the service simple. His catch phrase was "a better FTP". Alas, Grand Central Networks has gone the way of many a startup.

Alignment with business is where the industry is finally heading... In an effort to differentiate themselves in the inevitable arms race like spiral of the integration market, webMethods has sought and received SWIFT Certification. Banks looking to upgrade their SWIFT integration technology now have a reason to seriously consider webMethods. iWay Software, as the adapter company has been at the forefront of supporting industry protocols such as SWIFT, FIX, HIPAA, etc. and is pushing into insurance with ACORD. Yes, vendors will continue to add bells and whistles to their products but I expect them to become more tightly aligned with various industries and their standards. A tool that is already aware of your industry's standard processes and vocabulary is incredibly valuable.