The E-Business Software Weekly is a series profiling trends and developments in software and applications that support e-business, the Internet, and other electronic communication channels. Look for a new story each week in this space.
The Myths of Web Services
The idea of delivering software applications over the Internet--a notion embodied originally in the application service provider (ASP) concept and, more recently, in the Web services model--is one of the most provocative and enticing on the technological landscape. As Maria Atansanov writes in a recent issue of Ziff Davis' Smart Business, the strategic benefits of using an ASP can seem "irresistible to companies large and small. Why pay software companies every year for the privilege of installing and supporting their applications on an arsenal of PCs when you can hire out, accessing the same programs across the Net?
"Your company gets best-of-breed applications as soon as they're available," she goes on. In fact, "ASPs have reduced client companies' IT costs in some cases by as much as 20%. And you gain speed to market. What's more, you can focus on your core business, rather than on the Byzantine art of building your own IT shop."
Sounds like a great deal. And yet, Atansanov is merely setting up a straw man. Her article's title--"The ASP Trap"--gives away her story's punch line. She concludes scathingly: "Like many technology sectors, the ASP industry is in perpetual turmoil. And for far too many companies, it has become a pit of deception, exaggerated claims, and unfulfilled promises." Similarly, an August 2001 survey by the Associated Press reports that, despite the backing of many of the world's biggest software companies, "online software rentals seem like a pipe dream to many industry executives and analysts."
The software outsourcers themselves appear to be feeling the effects of the dimming perceptions of their model's viability. The Smart Business report notes this forecast by industry analysts: of the 1,800 current ASPs, just 40% will still be in business by the end of the year. "The phenomenon of 2001 will be the vultures circling," the magazine quotes Dave Boulanger of AMR Research as predicting.
There is abundant evidence for this claim. Notes the report: "Still unprofitable, [the leading ASPs] traded below $2 a share in early May, having lost more than 90% of their value year on year. KMPG dropped its stake in Qwest CyberSolution," a major ASP. "And software makers like J.D. Edwards, which had recently gotten into direct hosting, quietly retreated to selling software the old way."
Perhaps the idea of delivering software as a service--whether through a traditional ASP or its more modern offspring, a Web-shared service--really is nothing more than a pipe dream, and a very expensive, even dangerous one at that.
Ideas and Execution
To which proclamation a more prudent analyst would declare: "not so fast." Just as the dot-com shakeout of 2001 produced a loud but ultimately misguided cacophony of forecasts of the entire Internet's demise, the ongoing collapse of the traditional ASP sector, sobering though it is, should not be taken as the death knell for the concept of delivering business applications over the Internet. After all, it wasn't the Internet itself that doomed so much of the dot-com sector, just a lot of bad ideas and bad executions. As one keen observer of the past few years has remarked, investors "have funded all kinds of bad business models over the last 24 months. And a bad business model is a bad business model, whether it is dot-commed or not." Indeed, the only result of "dot-coming a bad business model is to make mistakes faster."
The same conclusion applies with equal force to the software-as-service concept as it does to dot-coms in general. Just because one implementation of this concept--the ASP model--has largely failed does not mean that the underlying idea itself is flawed. In fact, the success of numerous online banks, online brokerages, online travel reservation systems, and online auction systems proves that business-relevant applications can be delivered in a purely Web-based environment. Even many more basic business applications--like Salesforce.com's sales force automation and customer relationship management (CRM) applications--have been made available completely online, and are quietly transforming the delivery of these business-critical software functions.
So why the analysts' decidedly downbeat projections for the future of software outsourcing, and in particular for the delivery of business software applications as a service over the Internet? Setting aside the "piling on" mentality of pack journalism that has played a such a major role in the ups and downs of the technology sector in recent years, the most important explanation may lie in the analysts' failure to distinguish between the traditional ASP model and the newer Web-shared services concept. As other columns in this series have noted, the two technology models are fundamentally distinct. Relying on a traditional ASP is the technological equivalent of leasing a car from a third-party financing company: you pay a monthly fee and are spared much of the upfront purchase cost and ongoing maintenance, but you often don't save any money over the long-term and you wind up with pretty much the same car and the same capabilities as you would have had if you had bought the vehicle outright.
Web-shared services are a different entity entirely. Relying on a software-like application delivered as a Web service is the technological equivalent, not of leasing a car, but of leasing, say, a personal jet pack. As with leasing a car, Web services save a business customer much of the upfront cost and ongoing maintenance of purchasing and installing an application in-house. But the customer actually spends less--sometimes far less--over the long-term because the per-unit costs of delivering a Web service are usually appreciably less than delivering the same functionality in a traditional software or even ASP mode. And, what is more important, just as a personal jet pack frees the user from the rigid confines of a fixed road system, a purely Web-based service frees its users from the equally rigid confines of desktop-based software, delivering such benefits as easy anytime/anywhere access, collaborative computing, and real-time data updating that are simply not available in most traditional desktop or client/server systems.
The failure to make this key distinction is perhaps the most common reason why the deficiencies of the ASP model are often attributed, in one broad stroke, to the whole software outsourcing model and hence to the Web-shared services concept as well. Still, a number of potentially troublesome objections to the Web services model have been raised--call them the "myths of Web services"--and it is worth addressing them one-by-one.
Myth #1: The Vaporware Myth
The most basic myth of Web services is what one might call the "vaporware myth"--that is, that Web services are an illusion, that they don't really exist. For instance, in an otherwise favorable March 2001 article, IT World quotes one technology watcher as saying that "there are so few examples of Web services in action that it's difficult for developers to understand the value proposition." Similarly, in a glowing paean to the future of Web services, Phil Wainewright, the founder of ASPnews.com, nevertheless notes that "the only live, operational Web services we can point to so far to back up our claims are just a handful of cool gimmicks for Web sites."
Really? All of this would come as a surprise to Marc Benioff, the chairman of Salesforce.com. After 13 years at Oracle, including a stint heading the software giant's new technology development division, Benioff and several other colleagues set out on their own to validate Benioff's belief that one of the greatest new opportunities in technology lay in using the Internet "as a distribution vehicle for enterprise services, enabling customers to quickly sign on to a Web site and get all the features they'd get from a traditional application, but with much greater ease of use." The new venture, Salesforce.com, was launched in March 1999, and its enterprise subscription service went live with its first users in the fall of the same year. Although Salesforce initially focused on sales force automation, it has since branched out into customer relationship management (CRM) and marketing automation, and now offers nine integrated modules--all delivered directly and completely over the Internet.
Salesforce's pricing model is simple: a subscription to the service costs $65 per month per user. For that, customers receive not only powerful enterprise applications, but also freedom from the burden of having to maintain the applications' IT infrastructure in-house. Salesforce's offer has been persuasive. Already, more than 3,000 companies, representing some 50,000 users, have subscribed, making the firm one of the fastest growing CRM companies in the world. The growth trends are so positive that Salesforce confidently predicts profitability by early next year.
And Salesforce is hardly alone in the Web services arena. At least three companies are successfully delivering accounting services online. Intuit's Web-based tax preparation service is in its third year, and has been an unequivocal success (in fact, the IRS itself has spent millions of dollars on television commercials promoting its own online tax filing services). Several companies, including a number of industry leaders, either have or will soon launch online payroll services. Online expense reporting and professional time and billing services have become well-established. Online customer service applications, online bill presentment and payment, and online faxing have become standard fare at both online and offline businesses. Web-based teleconferencing, led by companies like WebEx, has been similarly integrated into day-to-day business operations. And all of this excludes such prominent early examples of Web services like those mentioned earlier: online banking, online brokerages, online travel reservations, and online auctions.
Web services an illusion? Not a ghost of truth to the idea.
Myth #2: The Customization Myth
A second common objection to the Web services idea is what one might call the "customization myth." The Associated Press, for example, reports that "the skeptics doubt large corporations will dependŠ on cookie-cutter online subscription services for software critical to their day-to-day work."
Really? Forget about the specific software delivery mechanism for the moment. How many corporations draft important press releases in Microsoft Word? How many do budgeting and forecasting in Microsoft Excel? How many corporate executives use Microsoft PowerPoint to prepare Board presentations? How many corporate graphics departments employ Adobe Photoshop and Quark XPress to design marketing publications? How many thousands of corporate Web banner ads have been created using Macromedia Flash? For that matter, how many Global 2000 companies use AT&T's telephone services or FedEx's shipping services or American Express' credit card services?
The point is, all of these common business functions are critical to the day-to-day operations of virtually every major corporation in the United States. Yet each piece of software and each portfolio of services is nevertheless "cookie cutter" to a substantial degree. While the software programs mentioned above, for instance, allow for some small amount of customization (as with the positioning and contents of the toolbar), they are defined by their rigorous standardization: all users have more or less the same tools and functions at their disposal. Similarly, while service companies like AT&T, FedEx, and American Express provide a multiplicity of classes of service, they generally don't allow much customization within a service class other than permitting users to select from a menu of add-on options (e.g., call-waiting, caller ID, and so forth).
As it happens, most Web services are or can be made every bit as "customizable" as these staples of the modern corporate environment. The notion that Web services must be more customizable than this--even to the point of being almost infinitely malleable--reflects a serious misunderstanding of the role of computer software in corporate operations. To be sure, some core corporate software applications, like ERP systems, supply-chain management tools, e-commerce platforms, and vertical-sector solutions, do demand a high degree of company specificity, and so are not yet well-suited to delivery via Web services. But the vast majority of software applications and services used in a corporate setting do not require extensive company-by-company customization, whether delivered via Web services, packaged software, or a non-software solution.
Myth #3: The Ownership Myth
The "ownership myth" is perhaps the most loudly proclaimed of all of the objections to the Web services concept--and probably the most specious. The Associated Press quotes PeopleSoft CEO Craig Conway as saying that "large companies are never going to trust someone else to run their software. If you believe technology is your most precious asset, you are not going to let go of it."
With all due respect to Craig Conway, that striking pronouncement is probably news to AT&T and the other telecommunications providers who own the vast majority of technology over which virtually all vital corporate voice communications take place. It's probably news to the companies that use sophisticated reproduction technology to print corporate annual reports and brochures and SEC filings. And it's probably news to the makers of office machines like copiers, the vast majority of which are leased, not owned, by their corporate users.
In fact, the principle of not owning the technology to deliver certain key business services is well-established. As Microsoft's Robert Hess notes, "if you're a company that wants to send a product to your customer, you call up FedEx or UPS or something like that, and they come pick up the product and take it away. You aren't forced to build a FedEx truck, and put a FedEx uniform on one of your employees, and drive around looking like a FedEx guy just to get the stuff shipped." Indeed, corporations already outsource a vast number of business functions--and the technology that underlies them--simply because it's more cost-efficient to do so.
The ownership myth is no less valid when applied to a company's nominally "core" technologies than it is when applied to day-to-day operational technologies and services. In an August 13 report on the emergence of Web services, the San Jose Mercury-News observes that "most corporate executives still prefer to own their software because they are under the misconception that all the money they pour into technology gives them an advantage over rivals." In reality, the paper quotes Joe Cardenas, chief information officer for San Rafael, Calif., software maker Autodesk, as saying, "there is little difference [among companies] because most companies buy the same fundamental applications for standard business transactions."
Probed more deeply, the ownership myth reflects a failure to distinguish between two very different dual meanings of the concept of "ownership." It is beyond question, for instance, that the ownership of unique technologies and processes contributes to significant competitive advantage in the marketplace. Wal-Mart sued several former employees a couple of years ago because of the risk that the employees could reveal the retailer's sophisticated inventory-management practices to a competitor--a key factor in Wal-Mart's low-price leadership in the consumer merchandise field. And Amazon's personalization and recommendation technologies clearly constitute a key advantage that the online bookseller holds over its rivals.
But most corporate technologies are neither unique to their corporate users nor pregnant with competitive advantage. Whether the technology in question was developed by SAP, BroadVision, Siebel Systems, J.D. Edwards, Sun, Oracle, or IBM, the ways in which the technologies are employed to support business processes and value--not the ownership of the technologies themselves--is what primarily determines their contribution to a company's competitive strength.
Myth #4: The Security Myth
A more emotionally laden aspect of the ownership question involves the ownership not of technology, but of data. Companies, it is said, will not use Web services or any other form of hosted technology solution because of their desire to keep their data "in-house," sitting on secure servers in, say, a concrete-reinforced corporate basement. Given the repeated instances of computer hacking and other network security breaches experienced in the United States in recent years, this concern is easy to understand. But it, too, is largely a red herring.
Typically, the IT networks of all but the largest corporations tend in fact to be less secure than those of the major hosting and data center companies that house the data of the major Web services providers. Visit Exodus, or Intel, or any of a dozen of the leaders in the field, and you see multi-level physical security systems--complete with "war room" monitoring stations, video cameras, biometric scanners, metal detectors, and armed guards--that are no less forbidding than those found at the CIA or the Pentagon. In addition, there are the dozens of less visible but no less advanced data and network security protections, like access-blocking firewalls, identify-verifying digital certificates, and 128-bit SSL encryption that, tests prove, is not capable of being decoded with current or foreseeable technology for literally trillions of years. Few ordinary corporations can afford to invest the millions of dollars required to create an equivalent security infrastructure.
The degree of security protection notwithstanding, there is little practical advantage to in-house hosting of data that a solid service-level agreement and privacy policy cannot enforce in a hosted-data situation. Moreover, major corporations regularly transmit sensitive data via channels that are markedly less secure than hosted-data environments, such as email, telephones, and the U.S. Postal Service. And consider that banks, investors, and accountants routinely maintain highly sensitive corporate data in their internal files--again, in a manner that is typically far less secure than that found in most high-end hosted-data environments. (It is also worth noting that some of the most publicized examples of hacking and other Internet security violations have occurred, not at Internet hosting companies, but at ultra-secure government installations or at major corporations that host their own data in-house.)
Security of data is surely an appropriate and urgent corporate concern. But its satisfaction lies not in rejecting a whole class of highly efficient computing solutions; rather, it demands a regular audit of the entire corporate security environment and the assurance that this security infrastructure is as immune as possible to unauthorized access, regardless of the manner in which the data is stored or transmitted.
Myth #5: The Viability Myth
A central message of the spate of recent software outsourcing critiques, like the Smart Business article cited earlier, is that outsourced application firms could easily go belly-up or otherwise fail to deliver on their promises, leaving their corporate clients in a very costly lurch. This is no mere theoretical possibility. Already, hundreds of ASPs have gone bankrupt, and hundreds more, analysts say, are likely to do so before the end of the year.
Even prior to an ASP's bankruptcy, a legion of other difficulties await. The magazine cites the experience of one unhappy customer: the ASP "never installed the system. We never received the hardware or softwareŠ" Another battered customer complained: "Our site was continually down. [The hosting company] believed it could run ColdFusion on a Sun platform when it works better on an NT box." And another: "They got you up very quickly for low cost. But does the [additional] cost every time you need to make a change in your site fit into the development plans?" And still another: "As [the ASP] grew, their capacity to meet customer needs was strained. The last six months were bumpy, and we'd go offline." The magazine adds: "Virtually all ASPs promise secure servers, bulletproof uptime, and 24-hour instant customer service. But often, a third party is responsible for making good on the sweet talk. In the worst scenarios, the ASP is little more than a middleman between you and a host of companies--including the software makers that write the applications."
These are, of course, legitimate and sometimes wrenching concerns, and corporate customers clearly need to be wary of the vendors they are dealing with. But attributing such ASP-specific viability problems to the software outsourcing concept in general--a "viability myth," as it were--is a little like condemning the safety of automobiles because airplanes (another mechanical form of transportation) sometimes crash. The fact is, the vast majority of ASP service and viability problems don't even apply to Web-shared services, no more than the characteristics of lift and wingspan apply to automobiles. For instance, Web services require no on-site installation, hardware, or software--they are immediately accessible online once a signup process is complete. Web services are typically priced (much like telephone and cable service) with a fixed monthly fee, with defined add-on fees for optional services--they are not IT service contracts that compel the customer to pay hundreds of dollars every time an IT outsourcer charged with managing and fixing internal IT networks steps into the customer's building. And Web services are generally operated by the software developers themselves--they are very rarely run by application-aggregating middlemen.
Without doubt, Web services--like any Web site--can experience downtime and outages, which is why a solid service-level agreement should be a part of every credible Web service's portfolio of offerings. And individual Web services, like any business, one day could go out of business. But so could a company's telephone service provider, its benefits management consultant, or its accountant. Even with powerful new technologies like Web services, nothing in business is guaranteed, making both due diligence and the development of contingency plans as vital to a company's success as the selection of the most capable or cost-efficient service provider.
Making Sense
Web-shared services are an exciting addition to the panoply of technology solutions to which corporations of the emerging decade will have access. It is true, of course, that Web services and the larger concept of software outsourcing, like the Internet itself, are still maturing, and will experience their share of stumbles and glitches during the years ahead--still one more reason for corporations considering them to evaluate each given Web service as carefully as they would any other e-business initiative or technology.
But such novelty is hardly an excuse for rejecting the concept out of hand. As John Hillman, the director of sales for energy giant Enron and an initial Web services skeptic, recently told the San Jose Mercury-News: "I think [with Web services] you are going to see a whole rebirth of the ASP model. It just makes too much sense for it not to work."
A great many corporate Web services customers, it appears, are already finding that to be the case.
ADVERTISING FEATURE
