Something about the way people are talking about web services seems eerily familiar. I'm forcefully reminded of my previous experiences with ASPs, an industry whose rapid boom and bust hurt a lot of entrepreneurs and investors. I warned of the dangers last year in an article called Web services claptrap. Here's some updated advice listing the main pitfalls web services vendors and solution providers need to watch out for, excerpted from my ASPnews column this week:
It's not a market, it's a mechanism: "There is no natural market for the technology alone."
It's not a quick fix for problem software: "Using it to plaster over structural problems in older software may seem to work for a while, but it just defers the day of reckoning."
Don't trust anything you hear from the big vendors: "IBM, Microsoft, Sun and the others ... are fighting a desperate battle to defend their existing market positions in the face of hugely unpredictable and disruptive forces for change."
SMB early adopters are not a mass market: "Most SMBs wait until they see new technologies endorsed by the big names in their sector."
Nobody's going to get rich quick: "You'll need many more customers than your traditional competitors before you get anywhere near breaking even."
There is no 'split' between J2EE and .Net the whole point of web services is to bridge the gap between the two. Tom Welsh explains why in a cogent opinion piece on The Register yesterday. I was amazed to hear recently from a leading manager at a big systems integrator that the biggest issue facing his firm's clients was consolidation, as if putting everything on a single platform will solve everyone's IT problems. Far better to concentrate on getting them to work in synch, I would have thought, and as Tom identifies, that's the real objective of web services. So if .NET does one job well and J2EE does others better, then you can deploy the right platform to do each job without landing yourself with a horrendous integration headache.
That's the theory, at least. In practice, there's a lot of ground to cover before it becomes palpable reality. Likening the task to NASA's quest to land a man on the moon, Tom concludes that "it will be a long job, because a set of wholly new technical problems needs to be solved and before they can be solved, they have to be identified." (Tom by the way is editor of Web Services Strategies, a newsletter published by Cutter Consortium. For its sixth issue, Tom says he has managed to assemble three "interesting and instructive case studies of how people are using web services in the real world.")
One of the barriers to interoperability is standards consensus or rather, the lack of it. Kendall Clark, writing at XML.com, deftly outlines the current status in Is There a Consensus Web Services Stack?, before getting sidelined into a discussion of the possible threat to open source vendors from patents. Here are some of his most telling conclusions from the first half of the article:
"The web services 'stack' is a semimythical creature, composed in equal parts of technical specification, running code, and marketing campaign ... even with a rough consensus, interoperation can be messy across different implementations of the web services stack; this is the case, in part, because people still aren't sure whether XML is a data model, just a syntax, or both."
Quoted in a CNET article on IT spending patterns, Krispy Kreme CIO Frank Hood makes the eminently sensible point that his company's priority is honing its core business, rather than acquiring the latest technology: "I don't think we need a killer app ... As CIO, I need to focus on business processes and the business of making doughnuts, not adding megabytes. Our killer application is our doughnut."
Interestingly, the comment echoes this sentiment from a recent Harvard Business School review of Howard Smith and Peter Fingar's book on next-generation business process management: "If you think Fordís most important product is the automobile, think again, say the authors. Instead, the process of making automobiles, business process management, is what really counts at the automaker."
In other words, the essence of a business is not its end product but how it produces that end product. Krispy Kreme's killer app is its ability to produce appealing, satisfying doughnuts at a competitive price, time after time. That's a business process (or more accurately, a collection of processes), and in many businesses, technology plays only a very small, supporting role in the killer processes that define the business.
The notion of the killer app arose in the early days of business computing. The spreadsheet was seen as the killer app that made the PC popular. Many early buyers wouldn't have cared whether it was a PC, a toaster or a bicycle, just so long as they could load up Lotus 1-2-3 and run their financial projections through it. What really mattered, though, was not the spreadsheet software itself, but what it enabled its users to do better, faster or more cheaply. The new technology allowed them to create new killer processes, which redefined their jobs and what they were able to achieve.
All along, the search for the next killer app in technology has been missing the point. Countless entrepreneurs, venture capital investors and corporations have been wasting their energies and their money trying to create software and machines that would somehow become the next killer app, when what they should have concentrated on was creating tools that help users unlock killer processes.
As the CNET survey shows, all those years of touting technology for its own sake has left buyers with jaded appetites. They've already bought enough applications and infrastructure to last them a lifetime, they feel, and yet all this investment has produced few if any new killer processes. Now they're thinking that maybe they should just get on with refining the processes they've already got, using the technology they already bought. This is bad news for people who sell technology, but it makes a whole lot of sense to people who run businesses.
posted by Phil Wainewright 2:26 AM (GMT) | comments | link
Tuesday, February 11, 2003
Happy 5th birthday, XML
An article over on CNET celebrates the fifth anniversary of the release of XML 1.0, but blots its copybook by perpetuating the notion that XML is a language: Developers reflect on the Web's lingua franca. As Propylon's Sean McGrath commented on my posting The language gap last week: "Perhaps the biggest curved ball thrown in this industry ever is that XML is a 'language'. Boy if I had a dime for every time I've explained why it isn't a 'language' in the way people normally interpret the phrase ..."
Five years doesn't sound very long, but my how things have come on since then. The CNET article quotes a number of developers who formed the original group that drew up XML for the W3C. A few sentences stood out for me:
Jon Bosak: "[XML] can play an essential role in freeing its users from the big-vendor hegemony that has ruled the computer industry for the last 50 years. The ability of user communities to develop their own data formats is a powerful force for freedom from vendor control."
Jean Paoli: "The extent of people's imagination is phenomenal and, for the first time, XML empowers them to express their ideas into documents that are semantically rich."
Ron Schmelzer: "The pervasiveness and widespread adoption of XML has in fact changed the economics of technology adoption--while in the past it might have been economically feasible to work in a vendor-proprietary, closed environment, the reverse is now the case."
Actually, Ron wasn't one of the original XML Working Group, but he and his ZapThink colleague Jason Bloomberg both get to contribute lengthy quotes to the article. ZapThink, an analyst firm that focusses entirely on XML and web services, has grown its revenues more than 800 percent in the past year. Ron cites this as possible "evidence of the growth and adoption of XML and its importance to enterprises," though a big chunk of it is probably more directly attributable to Ron and Jason's ability to come up with strong, bankable quotes for every journalist writing about the topic and of course to the remarkable scope and accuracy of their analysis. Being in the right place at the right time helps, but producing a quality product is the foundation of their success.
posted by Phil Wainewright 2:08 AM (GMT) | comments | link
Monday, February 10, 2003
New content and navigation
Loosely Coupled launches its first feature articles today, and we have tweaked the site navigation to coincide with the change. Our stories will examine topics of interest to enterprise adopters of web services and business process technology. We launch with three looking at EAI, ebXML and vendor training respectively and will be publishing at least one a week from now on.
There is a new RSS 2.0 feed dedicated to the stories section, which you are welcome to subscribe to if you use an RSS newsreader. But if you don't want to end up with multiple feeds from the site, bear with us for a week or so, when we hope to introduce a composite feed that will carry all the site's new content a true 'rich site summary' (or, to coin a phrase, 'really definitive feed').
We apologize if today's changes to the site navigation the first time we've changed the main navbar since the site went live almost a year ago causes any disruption to your habits as a regular visitor. (For the next hour or so you may find some pages still have the old navigation; that will quickly be fixed). Based on our analysis of how visitors use the site, we believe on balance the new layout will prove to be a great help to the majority, and a temporary inconvenience to just a small minority. However, as always, we appreciate your feedback, particularly since we have a more radical redesign currently in the works.
Like any dynamic composite application, the Loosely Coupled website will continue to evolve. You may notice further changes to some pages and services over the next week or two, and we look forward to introducing more new sections in the coming months. In the meantime, we hope you'll enjoy reading our new articles.
posted by Phil Wainewright 7:40 AM (GMT) | comments | link
Assembling on-demand services to automate business, commerce, and the sharing of knowledge