Web Services: Halfway Through the Integration Marathon
Monday, April 1, 2002
Web services will provide the same kinds of challenges and opportunities for business growth over the next five years that the commercialization of the Internet brought to the last five.

As Web services evolve, we will see the same development cycle that we have seen with every major technology platform: there’s the hype as the first movers declare themselves; excitement builds to gold rush frenzy as competitors rush in; checkbooks are whipped out to finance every cloned entity. What follows is a sweeping set of failures and, finally, there is success — for a few.

The trick is to be among the few.

Web services winners — companies that combine true value for customers with a sustainable business model for founders and stakeholders — will be those whose leaders are willing to stay behind the curve, gathering intelligence, building slowly and methodically, while staying nimble enough to make their move when the time is right. Preparation is everything. Patience is key. Luck will play a big part.

It’s a lot like speed skating — Apolo Anton Ohno style. His method was to ride behind the early sprinters, watching carefully for his chance, before making the last smooth inside move to go for the lead. Of course, it didn’t always work as he planned. He lost a few medals through falls, bad timing, and disqualifications. Of course, more often, he won the same way.

Entrepreneurs looking at the opportunities and challenges that Web services offer need to take Ohno’s example to heart. Recognize that we’ve barely begun, prepare carefully; work hard and intelligently (stay agile, expect setbacks) while always keeping the goal firmly in mind.

The big question is ‘To what end?’ Answer that, and you’ll have a business.
Otherwise, the customers’ checkbooks will remain closed. Too many companies’ CEO’s, CIO’s and CFO’s are refusing to buy (or buy into) the latest, greatest solutions because they have been burned so often in the past (According to Forrester Research, financial institutions alone spend 40 percent of their IT budgets trying to tie their existing applications together). They’ve been burned by “security”, by “seamless integration”, by “robustness” — by all the optimistic promises and key marketing words used by virtually every enterprise software or services company.

ABCs of EAI: Lessons Learned
First, it is worth reflecting for a few minutes on Web services predecessor — Enterprise Application Integration. Over the past three to four years, the development (or more precisely promised development) of EAI commanded the same attention and fascination now focusing on Web services.

EAI promised to unite myriad applications into a dream of seamless efficiency. It predicted a B2B world where all members of the extended enterprise, suppliers, customers, and partners would skip lightly across networks, accessing diverse applications, sharing data, and building, as they did so, the new world of networked business processes.

Good idea. Bad implementation.

A few years and billions of dollars later, EAI looks more like “Fear and Loathing on the Network.” Fundamentally hindered by the development of multiple solutions from multiple developers, without a common standard, EAI solutions have become expensive, cumbersome, and jury-rigged stopgaps on the road to Web services.

EAI required building point-to-point solutions, creating a tightly coupled world. Web services force a new perspective: from now on solutions will have to work effortlessly in a loosely coupled world. Today, companies still trying to make EAI work are demanding to know their suppliers’ strategies for integrating their solutions with Web services.

We’ve Only Just Begun
Today we are only at the first stage of the Web services cycle. But we are already on a much better path than the one taken by the EAI solution providers of the past few years.

We have the beginnings of standards that facilitate interoperability. Although some developers may argue the standards are klutzy, they are here to stay.

Extensible Markup Language (XML) won industry endorsement in 1998. Simple Object Access Protocol (SOAP) was introduced at the same time, followed by Web Services Description Language (WSDL) and Universal Description, Discovery, and Integration (UDDI), which first shipped less than a year ago, in July 2001.

The next step is to define business processes. With the February 2002 foundation of the World Web Services Interoperability Organization (WS-I) backed by Microsoft, IBM, BEA and about 50 other major players, it appears likely that IBM’s WSFL and Microsoft’s XLANG, used by its BizTalk server, will join to win acceptance as the first business process standard, although EbXML is also in the running.

And then comes the hard part: the standards, protocols and processes now in development only address the (relatively) simple problems. The big issue — and arguably one that will prove the most lucrative — will be overcoming interoperability challenges with security, authorization, authentication, and routing protocols that are both reliable and scalable.

Web services are the logical next step — and big leap forward — towards realizing the dream of true distributed computing. Put simply, Web services provide a new way of connecting and messaging. But preparing to move into the Web services arena first requires an understanding of what and where Web services are now and where they are going.

The Internet is driving the Web services opportunity. It was just over a decade ago, when the National Science Foundation approved the commercial use of the Internet, that the Worldwide Web was officially born. Four years ago, there were 1.7 million Web sites. Today, there are more than 36 million and 110 million domain names have been registered.

Despite the incredible growth of the Internet and related intranets, the level of technical interoperability on the Web is immature. Web services addresses that problem: the goal is to provide a set of standards (EAI’s weakness) that will allow all kinds of applications from disparate hardware platforms and operating systems to interoperate seamlessly and efficiently. The business opportunity will be combining applications to produce new solutions and services that provide significantly increased value to companies.

However, solutions are also needed that work with legacy systems and make them work better — as Alfred S. Chuang, Founder, President, and Chief Executive Officer, BEA Systems pointed out at siliconindia’s February 2 Tech Factory West conference, 90 percent of companies are still running legacy applications. Companies need to develop ways to facilitate straightforward integration that adds value at every step along the way.

Therefore, it’s important to note that the first markets will be for Web services application interoperability on intranets. Companies have invested heavily in developing their own Virtual Private Networks (VPNs), but cost, accessibility, reliability and scalability within those VPNs are still thorny subjects. Those technology interoperability issues will have to be solved and proven rock-solid inside firewalls before any company will consider implementing solutions outside the four walls of the company.

Work Hard and Smart
Right now, it looks as if the timeframe to the most fundamental Web services capabilities is anywhere from 18 months to three years down the road. The technology will develop in ways that we cannot foresee and competitors will appear.
Any company wanting to be a player in the Web services arena has to ensure that its applications and business models are programmed to work with the standards that are evolving.

Any supplier wanting to move ahead with the crowd should be working now on adapting these standards to existing products, and including them in their plans for future products.

Any customer seeking interoperability needs to ensure that its suppliers are part of this mainstream movement. As Web services evolve, expensive and time-consuming solutions that are tailored to support individual protocols will become a thing of the past.

Those criteria need to be part of every new and revised business plan. That’s the low-hanging fruit. The bigger question is how an entrepreneur can start now to develop a company that will deliver the technologies, applications and services that will provide real value when Web services happen.

Keep the Goal in Mind
As Web services develop, there will be early leaders, slips and falls, and a lot of juggling for position. There will be choke points that no one will easily figure out how to break. Success will not come easily. It may not come exactly as expected. Ohno didn’t get five gold medals, but he certainly emerged a hero. In his winning races, from his position towards the back of the back, he saw and clearly evaluated the blocks before making his move.

The choke points that will appear as Web services grow, will provide enormous opportunities for companies that have maintained a similar steady pace behind the leadership curve.

An obvious choke point will be network traffic. It’s one thing to have one small stock quote whipping in and out of applications and updating stock portfolios; it will be quite another challenge when whole business processes, and every piece of them, are finding and sharing data, at the same time, on the same network. Security will be a huge issue. So will reliability.

The breakthrough killer app will start with a Web service for universal customer management. “All” that’s needed is to get the data that lives in so many different repositories and that is manipulated by so many different applications, into a standardized customer directory.

Creating a unified customer reference that is secure, linked to, and accessible from, all applications across the network is a daunting challenge. Once solved, it will quickly expand outside companies’ VPNs and beyond their firewalls to provide the basis for ubiquitous supply chain integration.

Addressing those issues first behind company firewalls is akin to assiduously training before the big race.

Which gets back to the original premise: taking advantage of the new opportunities and challenges Web services promise will require patience, hard work, a solid business plan, agility, commitment, and a good dose of luck. The winners, who will emerge from the back of the pack, will have gained invaluable experience and established their credibility through a series of wins in smaller arenas.

When the finals are underway, and the opportunity opens, they’ll be the ones ready and able to “go for the gold.”

Mark Kvamme is a partner with Sequoia Capital, an early stage investor in Apple Computer, Cisco Systems, Network Appliance, Oracle, PMC Sierra, Vitesse Semiconductor, Yahoo! and many others. He is currently a director of Bang Networks, Tonos, Sawyer Media Harmonic and Macromedia (MACR). Prior to Sequoia Capital, he was Chairman of USWeb/CKS and Chairman and CEO of CKS Group prior to the merger with USWeb. He has a B.A. in French Economics and Literature from the University of California at Berkeley.

Share on LinkedIn

Previous Magazine Editions