Sunday, January 07, 2007

The SOA Forecast for 2007

Document ID: ZAPFLASH-200713 | Document Type: ZapFlash
By Ronald Schmelzer
Posted: Jan. 03, 2007

For those in the United States, 2007 has started out with a bang – two major snow storms in the Rocky Mountains and a persistent warm front that’s keeping the entire East Coast in unusually warm weather for this time of year. Even more so, it’s football season in the US and both the New England Patriots and Baltimore Ravens are in the playoffs – the two teams representing ZapThink’s two office locations. Of course, speaking metaphorically, with regards to SOA, 2006 ended with a bang and 2007 is already showing considerable warmth and competitive vigor. ZapThink has seen SOA take off even more aggressively than we anticipated at the beginning of 2006, and all indications show that SOA strength will be further reinforced and expanded in 2007 to many corners of the IT environment, throughout the world and in many different industries. And so, during this season of sultry winter weather and competition, it is time to evaluate our predictions from the previous year and forecast th e architectural and competitive climate for SOA in 2007.
2006 SOA Predictions: Four for Four!
In last year’s 2005 SOA Scorecard and 2006 Predictions ZapFlash, we predicted that companies would move beyond their tentative SOA projects to more substantial projects leveraging emerging best practices for all aspects of the Service lifecycle. More specifically, we bemoaned the fact that companies were thinking too much about having a bunch of Services (so-called “ABOS”) and not enough about an architecture oriented towards the ongoing evolution of those Services. We surmised that companies would soon wake up to the fact that SOA demanded true enterprise architecture activities, artifacts, and skills, and less-so just rote software development and technology implementation. Indeed, we believed that one side-effect of the maturation of SOA projects was that companies would finally give proper emphasis to the use of a registry/repository throughout the Service lifecycle, effective Service lifecycle manage ment, and a comprehensive IT governance framework. We predicted that an emerging set of next generation SOA projects would put some enterprises in technology leadership positions, and indeed, it seems that all of those predictions came to pass in the year preceding.

Continuing on the thread discussed earlier with regards to maturing SOA implementations, we also predicted the SOA governance space was ready for a major shake-out and consolidation. Boy, were we right on target! 2006 bore witness to the frenzied acquisition of all things governance and registry related as well as the emergence of Service lifecycle vendors as the next must-have technology in any SOA infrastructure portfolio.

Our next prediction was that emphasis would shift from being solely Service provider centric to an increase in awareness, and implementation, of Service consumer-centric tooling and methodologies. For sure, 2006 emerged as the year that Ajax moved from being an acronym to being a metaphor to all things Service consumption-centric that enhanced the browser experience as a first-class environment for user interaction. But even more so, 2006 saw the emergence of real Service consumption and composition tools that aimed to facilitate the creation of Service-oriented Business Applications (SOBAs) and Enterprise Mashups that free the consumer-focused developer from having to worry about the location and implementation details of the Services they consume. Indeed, have we already reached the Web Services Tipping Point?

Another prediction we made last year was that regulatory compliance would take center stage in 2006. Specifically, we said that there would be a dramatic increase in the number and capability of SOA-based compliance solutions coming to market in 2006. Some of these solutions may be in the form of composite applications, while others may simply be sets of Services. Either way, SOA will become increasingly critical to many enterprise's compliance initiatives. The result was more compliance-focused SOA initiatives than we have fingers on our hands to count! Multi-national firms in particular found that compliance with global regulations that continue to change at relatively unpredictable rates has, as a result, an unpredictable burden on their IT departments and thus they must find a way to grapple with such external change factors. ZapThink released a survey of SOA consulting firms in mid 2006 with the main finding th at SOA itself is a hard sell for most companies, but a focus on selling solutions to common business problems, of which solving compliance problems is a key one, is a cake-walk.

2007 Prediction #1: SOA Quality and Testing Market the Next Must-Have Space
So, it seems that we were right on target with our 2006 predictions. Maybe they just were too obvious when we made them at the end of 2005. So in this iteration, we’ll try to make some predictions that are less so, or at the very least might be up for debate. One prediction is that the new, hot area in the SOA marketplace are testing and quality tools and solutions for SOA. While SOA-focused testing and quality vendors have been around for at least five years (we first covered them in our Web Services Testing report in 2002), the impetus for widespread adoption of testing across the full Service lifecycle hasn’t been there for much of that time.

However, that is set to change in a big way in 2007. Now that companies are through with their proof-of-concept and play phases of Web Services-centric adoption (less SOA and more ABOS), they have to grapple with the real issues of evolving SOA: change and version management, testing in a production environment (runtime testing), quality assurance of all SOA artifacts including contracts, policies, schema, models, declarative composition and process metadata, and other forms of metadata, as well as deal with change-time and run-time governance that demands more than simply testing the code that exposes a Web Services interface. As a result of this heightened awareness of the real challenges in maintaining a SOA implementation, demand for SOA quality and testing solutions will skyrocket in 2007, leading to greater acquisitions, increased consolidation, new venture creation, and boatloads of case studies on the topic. Watch this space for the real action.

2007 Prediction #2: Enterprise Architect Drought
One dire prediction for 2007 is that there simply won’t be enough qualified and SOA experienced enterprise architects (EA) around. Sure, there might be many “paper architects” – that is, those that claim EA and/or SOA skills on their resumes or in their career history, but much of that experience will be attendance at a few vendor-heavy SOA courses, the development of Web Services-centric interfaces, and a sore lack of any methodology, modeling, Service lifecycle, or governance experience to speak of. Yes, it might just be that the biggest force gating widespread adoption of SOA is not the technical complexity of SOA projects (one can actually say that the technology part is relatively trivial), but rather the organizational, architectural, and skill gap that most companies have in making this architectural change a reality.

ZapThink has seen first-hand evidence of this lack of EA skills. First, there is a significant demand in the marketplace for experienced SOA talent. Second, we are seeing a burgeoning of SOA consulting companies that offer kick-start approaches to SOA in which they supply the experienced architects and their customers supply the heavy-lift labor to implement the Services. Already we’re starting to see a bifurcation in the IT community between architect and developer, with development seen as an increasing commodity whereas architecture is an increasing scarcity.

ZapThink is doing something about this lack of EA talent. In fact, we’re doing two big things about it this year. First, ZapThink launched its Licensed ZapThink Architect (LZA) program in late 2006 with little fanfare, but we are planning to trumpet its increasing value and successes with greater volume throughout 2007 and beyond. While certifications do exist for aspects of SOA, they are usually specific to a particular technology du-jour or an individual vendor implementation. Increasingly, individuals are looking to go beyond understanding of what SOA is and dive much deeper into how to do it right. The demand is thus more than just training, but also the backing of a qualified third-party organization to endorse their existing SOA skills, enable continuous improvement, enhance networking in the community, and enhance their current SOA-enabled careers. ZapThink is filling the unmet need for knowledge and credentials in th is area with its new LZA program. You’ll see increasing references to the LZA program throughout 2007, and you can dive deeper into details at

In addition to the LZA program, ZapThink will release within the next week or two its Architect Resource Center (ARC), a partnership with recruiting firm Excel Partner to list qualified EA and SOA resources on the ZapThink site to meet the needs for SOA resources. Indeed, ZapThink is now in the position to directly help end-user firms hungry for SOA talent, through our new ARC offering, and if our prediction comes to pass, we’ll probably find more demand for such resources than even we have the capacity to supply.

2007 Prediction #3: The SOA Suite Busts a Few Buttons… Open Source the Remedy?
The final prediction ZapThink will stick its proverbial neck out to make for 2007 is that the SOA suites might get a bit too big in 2007. Warren Buffet is fond of saying, “don’t ask a barber if you need a haircut”. What he means to say is that vendors of a service or a product have a self-interest in promoting the need for products that probably won’t keep you away from their sales department for long. While there’s no doubt that SOA vendors are solving real problems with products that, in many cases, actually work, the problem is that these SOA solutions have grown from individual, point-products that solve discrete SOA problems to gigantic SOA suites that seem to be every bit the monolithic platform that they had been intended to replace. In fact, we have been quoted as saying that monolithic SOA platforms, in which you have to buy all the various parts of the solution in order to get the true value as promised by the vendor, is at its core an un-SOA philosophy.

The idea of SOA is to promote loose coupling and composition in an environment of continuous change and heterogeneity. A proper architecture should be blind to the underlying runtime infrastructure, messaging protocol, or application server environment. Want to implement SOA on a Commodore 64 and COBOL-based mainframe environment using CORBA over a token-ring network? Knock your socks off – there’s nothing un-SOA about that. But saying that you need to use a single messaging bus, Service container environment, or management infrastructure to get any value out of SOA is being disingenuous to the SOA mindset, and a disservice to end-users looking to finally take control of their own IT environment.

One potential antidote to the vendor indigestion that many end-user firms are experiencing is the burgeoning of open source efforts that take a laissez-faire approach to technology implementations. ZapThink is starting to see increasing use of Apache and a multitude of other open source technology underpinnings for a variety of the runtime and design-time aspects of SOA, including Service exposure, messaging, composition, rich client, and even security and reliability offerings. The Eclipse Foundation has really hit a stride with regards to dominating the development landscape for many SOA efforts. While it would be hard for us to say that such open source efforts would crimp the businesses of the well-established software vendors, it is clear that those efforts are not outlying or rogue efforts proposed by small departments of large firms or by small companies. Rather, it is precisely the behavior of many large software companies that is encouraging or forcing the han d of large companies to seriously consider open source Pepto-Bismol for their vendor-induced heartburn.

The ZapThink Take
In addition to the above predictions, one significant observation in 2006 is the state of global adoption of SOA. Is the United States behind the rest of the world with regards to SOA adoption and maturity? Probably not, but it certainly isn’t ahead. Indeed, most of the largest, advanced, and oft-quoted examples of SOA projects in both the press and by software vendors in 2006 were in Europe, Canada, Asia, or Australia. There’s so much to analyze and understand in the trends towards global SOA adoption that we’ll leave this topic for a future ZapFlash, but rest assured this IT movement is unlike most others with respect to adoption.

There are probably a few more predictions we could make for the year, but we’ll be brief for the sake of your reading time and also so that we have more material for future ZapFlashes. Regardless of where you think SOA is heading, it’s clear we’ve crossed some sort of chasm with regards to SOA adoption. More companies than ever are already supposing that they’re going to Service-orient their systems and businesses , and this is resulting the growth and value of SOA throughout the world. ZapThink wishes you all a wonderful, healthy, and rewarding 2007!

No comments: