Over the past five years, companies have purchased a lot of software. First, there were concerns about Y2K preparations. Then, companies rushed to become “e-business ready.” In retrospect, both of these previous waves of IT investment may be better seen as aberrations rather than the norm. Combined with depressing news about the macro-economic environment, the significant slowdown in spending has made the technology sector a tough place to be.

Nonetheless, enterprises will be buying software in the years ahead. IDC and Robertson Stephens estimate that the application server market will grow 62.6% CAGR between 1999 and 2004 (to $11.3B total), while the total software infrastructure market will grow 51.5% CAGR ($18.9B total). Growth projections tell only one part of the story, however, and this article aims to provide a starting point for filling in the details.

Three Phases of Web Services Adoption

The Stencil Group has developed a model for the growth of this market that suggests web services will be adopted in three natural phases. Our perspective and the feedback from customers is that adoption will follow a fairly direct progression:

Phase 1, already underway, will be focused on internal integration and driven by cost-efficiency and re-use opportunities of existing assets, as companies begin to take advantage of the web services technologies break down the monolithic applications into much more discrete components. This phase is mostly one of ad hoc adoption.

Phase 2 will be characterized by a push towards making the transactions with trading partners to become more efficient. Flexibility and control are the key goals in this phase. Although some opportunities for establishing new revenue streams will exist, given that Microsoft Excel remains the most important “supply-chain management application” in the enterprise today, corporations are anxious to digitize and automate much of this data gathering and event tracking. The primary advancement from Phase 1 is a shift towards systematic strategies for developing, publishing and consuming web services.

Phase 3 will be about truly dynamic business relationships that are formed at trade or run-time. We believe this utopian picture is, frankly, more evangelism than reality and is five years or more in the future. Of course, some early adopter industries like financial services, telecom, and hi-tech manufacturing may be more receptive to this vision than other sectors will be.

Why Will Web Services Succeed, and Why Are Customers Excited?

First, distributed computing is a problem that engineers have been trying to solve for years. The emergence of some common standards, businesses placing an increased value on standards—both vendors and adopters—and a simpler solution than previous efforts will make the web services model work. For example, compared to CORBA, web services are a minimalist’s dream. CORBA tried to boil the ocean and was bogged down by enormously complex standards that were tightly coupled to specific instances of software code. Web services live by the K.I.S.S. motto (“keep it simple, stupid”) and focus on moving data around using XML to keep inter-application communication very loosely coupled.

Enterprises are looking to apply web services or similar service-oriented architectures to specific business problems. Themes that we hear consistently are business control, a shortage of development talent, and the need for common network infrastructure. Given the increased competitive environment, companies are looking to operate more efficiently and increase their ability to operate in a real-time (or near-time) environment. The lack of skilled developers presents a problem to corporations that can be addressed by some of the very impressive tools coming out of Microsoft (for example, the VisualStudio.NET IDE is quite impressive) and other simple “SOAP wrapper” products from independent companies like Cape Clear. Lastly, we’ve heard the idea that “existing infrastructure plus new standards creates a new infrastructure.” On that note, some companies may be able to deliver a solution that help address the limitations (or high costs) of existing networks like EDI VANs.

We’ve suggested that software spending by enterprise IT groups is driven by three broad criteria: the need for sound investment, technical interoperability, and business process integration. Similar to any other investment, companies want to see a return on their money. For too many years, software has been a depreciating asset, not one that delivers an ROI. This must change. New software purchases do not happen in a vacuum. Therefore, new purchases must be certain to interoperate with existing infrastructure and applications, whether these are based on Java technologies, Windows applications, or homegrown code. Web services promise to make the “simple” act of connecting different pieces of software significantly easier. Lastly, as companies are looking to increase control of their business, the need exists for various business units to be connected and not operate as individual silos. The more granular nature of web services solutions will increase companies’ ability to do so.

Companies to Watch

This wave of software innovation is distinct in several key ways from earlier generations. The most significant difference from, say, the application server or client/server waves, is that the major vendors are involved from the start. What are they doing? IBM and Microsoft have been critical in creating and driving the standards, especially SOAP, UDDI and WSDL. Not wanting to miss the train, Sun has made a lot of noise, but hasn’t yet indicated a clear direction. After a slow start, BEA has joined IBM and Microsoft as a major platform player to watch. Beyond this group, the other vendors are worth watching (e.g. HP, Oracle, etc.), but we don’t believe that these companies will lead the market.

Early reviews suggest that the major software vendors will win the lion’s share of the market, and our first impressions concur with these conclusions. However, we haven’t forgotten that this market is at a very early stage. As the market coalesces, the picture will become clearer, but we do believe that a handful of start-ups will emerge as either successful independent companies or as very promising acquisition targets.

Of the forty plus private companies with whom we have met, there are a few who have impressed us. Be sure to take a look at Cape Clear, Epicentric (an iMinds portfolio company), Flamenco, Grand Central, Infravio, Interkeel, Invertica, Killdara, and ServiceMesh, among others.

Bottom Line

The emerging standards that comprise web services will succeed and, indeed, are already being adopted today. The market will grow in three discrete phases as businesses implement web services to increase their flexibility, improve business control, and make the enterprise more efficient as a whole. Still, one big question remains: will anyone beyond the major platform vendors make any money in this space? We’ll try to answer this question and more in the weeks and months ahead. To stay in the loop, I encourage you to subscribe to our web services newsletter.

Bill Robins
Principal
The Stencil Group

(Author's Note: This essay is an overview of many pieces that we have developed as part of this effort. A complete archive of our publicly accessible work can be found on our web site).