Printer Friendly

The next frontier: service automation.

This issue looks at what is arguably the most challenging question that software companies face in the 1990s: Where will they find new users? We see troubling signs that the industry's core markets, office automation and consumer applications, are reaching near-saturation levels. So far, high growth in international markets and the recent feeding frenzy in the Windows market have provided a reprieve--but what happens when the growth curves in these markets also begin to flatten? One hope, of course, is that next-generation technologies--pen-based systems, multimedia, groupware, hypertext, virtual reality, etc.--will attract new users. It's reasonable to expect that these technologies will generate some incremental growth within the current installed base, but we're nevertheless skeptical that talking spreadsheets or infra-red meeting planners will lure many brand-new users into the PC marketplace.

Instead, we now believe the next big round of growth will come from a market that is still almost invisible: service automation. The premise we offer in the following pages is that there are literally millions of service workers--in retail stores, restaurants, hospitals, customer service departments, travel agencies, field service organizations, consulting firms, and the like--who are about to move into the mainstream of the personal computer revolution. As these workers make the difficult transition from proprietary hardware and software platforms to PC-based environments, we predict they will generate a wave of new applications, new entrepreneurial opportunities, and new market leaders.


As economists and politicians keep reminding us, the U.S. economy is increasingly service-driven. Skim the pages of Fortune's annual "Service 50011 issue and this point becomes much less abstract: The corporate world is full of giant companies in areas like retail, communications, banking, and transportation that are chiefly sellers of services, not of tangible products. At the other end of the size spectrum, the small business market is dominated by retailers, restaurants, consultants, real estate firms, insurance agencies, and other service businesses.

In addition, most good-sized companies, regardless of their primary business, now have internal service organizations that may account for a large percentage of total headcount. (In the case of software companies, for example, tech support and customer service employees typically represent about 20% of total staffing.)

All of these various service organizations share at least one common characteristic: They invest heavily in service automation technology. Even small, undercapitalized retail shops almost always own at least one electronic cash register. (In fact, the installed base of electronic cash registers and networked point-of-purchase terminals in the retail segment alone currently exceeds 20 million machines.) Banks, mail order houses, supermarkets, travel agencies, and brokerage firms often end up spending more than the equivalent of a year's salary per employee on terminals and workstations.

Historically, almost all of this investment in automation technology has been spent on proprietary hardware platforms--machines based on custom processors, running non-standard operating systems and machine-specific applications. These systems typically cost two to three times as much as comparable PC-based systems, and (by PC standards) usually have serious shortcomings in such areas as interoperability, analytical features, and end-user programming capabilities.

But the service automation market is changing. within the last two years, PC-based systems have begun to edge out older, proprietary systems in almost every segment of the service automation market-- sometimes, in large organizations, thousands of machines at a time. The penetration of PCs is moving much faster in some segments than in others, but the overall trend is absolutely clear: Within the next decade, PCs will almost certainly become the dominant computing platforms in virtually all major service automation markets.

How fast will this transition occur? That's a tough question; if there's any trustworthy research, we haven't found it. Moreover, current PC-based systems still have some limitations. Important devices like cash drawers and bar-code scanners aren't well supported by office-oriented operating systems like DOS and OS/2, development tools and applications don't always exist, and distribution channels are relatively primitive. Moreover, we suspect that a fair number of customers will be reluctant to give up the security blanket of having one proprietary vendor take responsibility for their entire information systems package.

But even if we can't measure the absolute velocity of the service market's migration to PCs, there's no question that the pace is rapid and bound to pick up speed soon. The biggest vendors of service automation hardware, notably IBM and NCR, already offer PC-based platforms, and smaller hardware companies are jumping on the bandwagon because they see no real alternative. The result should be an immense installed base of tens of millions of new PCs and new users--an installed base that we believe ultimately will rival (and perhaps exceed) the potential of Windows as a new software market.


It's been said, only somewhat tongue-in-cheek, that the Big Four productivity applications for desktop PC users are "word processors, word processors, spreadsheets, and word processors." Desktop PC users tend to produce documents; service workers, by contrast, deal primarily with one-on-one customer transactions--answering questions often over the phone), processing sales, interviewing, making repairs, and the like. The service automation mainstream is thus defined by applications that directly enhance transaction efficiency and productivity.

Moreover, "productivity" is not a fuzzy concept to the people who manage service organizations. Automation solutions are judged explicitly by how they increase output per employee (measured in terms of transaction volume, sales, inquiries, etc.) without sacrificing service quality. A travel agent is expected to book more flights with an airline reservation system, a supermarket clerk should ring up more groceries with a bar code scanner, and--closer to home--a software support technician should field more calls with an on-line knowledgebase.

Most of the classic examples of service automation products are tied closely to specific industries, but these examples help define the general application categories that are likely to dominate future service automation markets:

* Transaction support: A large part of most "service" transactions is actually mechanical work: looking up and recording data about the customer or the transaction itself. Some service workers-- bank tellers, retail clerks, telephone operators--do very little except handle mechanical transactions. But even restaurant waiters, support technicians, medical professionals, and auto mechanics (all of whom provide more personalized kinds of services) spend a good chunk of time retrieving and recording information; in fact, one recent AMA study shows that doctors devote almost 40% of their time to writing up patient charts.

* Not surprisingly, service organizations tend to embrace automation solutions that shorten the amount of time spent on mechanical transactions. (Banks have pushed transaction automation especially far with ATM cash dispensing machines, leaving human tellers to handle less-routine customer transactions.) Customer service databases, bar code scanners, electronic forms, and point-of-purchase systems are especially relevant to a broad range of service environments. Our guess is that accounting software companies will take the lead in most of these niches; in fact, Great American, Dac, and State of the Art have all just rolled out point-of-purchase modules that are integrated with the rest of their PC-based accounting lines.

* Knowledgebases and expert systems: Service employees are frequently knowledge brokers": They answer questions about products, troubleshoot, and make judgment calls based on knowledge of corporate procedures. To handle these functions, service workers often have to spend a substantial amount of their time on essentially non-productive training and research tasks. Large service organizations--for example, American Express and Fidelity Investments--have begun to address this problem by giving hundreds of their service representatives access to on-line knowledgebases and expert systems. In some cases, customers who want information are able to navigate the corporate knowledgebase themselves, further reducing the need for human intermediaries. Borland, Lotus, and other software companies have found that 3%-4% of their tech support calls can be offloaded to an automated fax-back system. Similarly, Fidelity recently began offering its brokerage customers a PC-based product from MECA Software that retrieves stock quotes and executes trades directly through Fidelity's mainframe computers, with no broker involved.

* Scheduling and forecasting: Service organizations rarely have much control over their workflow: When a customer walks in or calls, it's assumed that the retailer, restaurant, or tech support group will have enough people on hand to provide reasonably prompt service. At the same time, when demand is low, it's equally important not to have lots of expensive employees doing nothing. Thus, a critical problem for service organizations is to be able to forecast fluctuations in demand (for a Christmas selling season, a new product release, or a Friday happy hour) and then to schedule employees appropriately to match these levels of demand. The productivity payoffs from automated scheduling turn out to be surprisingly high: for example, hospitals, retailers, and customer service groups have been able to reduce payroll costs by as much as 10%-30% simply by computerizing the labor scheduling process.

* Transaction analysis and modeling: Besides the problem of forecasting demand, most service organizations ultimately will need better tools for analyzing such issues as price elasticity, seasonal trends, and profit and cost ratios. Right now, most proprietary systems offer analytical features that are inflexible and poorly designed, and it's often impossible to transfer raw transaction data to third-party spreadsheets or graphing programs. (Sometimes the system's only output is a paper cash register tape.) Once enough service organizations make the transition to open PC-based systems, however, users will suddenly gain access to huge volumes of raw transaction data. As users begin to sift through this data to find ways to fine tune products and operations, it's likely that a robust market for analytical and modeling software will emerge.


For the moment, the service automation market is highly fragmented. Distribution channels are dominated by "vertical market" developers and hardware vendors; the typical customer is strikingly unsophisticated and risk-averse. PC software companies have always been wary about entering such markets, because of the high cost of sales and low unit volume. We believe that eventually mass-market opportunities will open up, but it seems clear that the early winners in service automation will have to do more than just get shelf space at Egghead. So what will it take to establish a beachhead in these markets? We've seen at least two good models:

* The direct sales model: Rosh Intelligent Systems (Needham, Mass.) is a venture-funded developer of knowledgebase development tools for field service applications. We met recently with Mike Levinger, Rosh's marketing vice president, who explained that his company sees itself in a direct sale, large-deal business." Rosh currently has about 30 large corporate and government customers--organizations with at least 300 field service engineers, who typically support complicated or fast-changing product lines. For these blue chip" buyers, Rosh provides extensive hand-holding, customization, and implementation help; 30%-50% of the price of a typical sale represents services, says Levinger. A key ingredient in Rosh's direct sales model, he adds, is a pricing structure that is geared toward corporate adoptions rather than single-unit, shrink-wrap sales. We're seeing customers now who are willing to make half-million dollar commitments to us."

* The alliance model: MECA Software (Fairfield, Conn.) recently developed a $120 electronic access program, Fidelity On-line Xpress, that Fidelity Investments markets to its brokerage customers. By creating a product for the customers of a single service organization, says MECA president Dan Schley, his company was able to open up a large special-interest market that it couldn't otherwise reach. MECA and Fidelity have been "absolutely flooded" with sales, he adds: In the first 30 days, we got 22,000 requests for information about the product." But Schley points out that Fidelity's enthusiasm for its alliance with MECA wasn't based on possible software or connect-time revenues. Instead, Fidelity saw On-line Xpress as a way to boost its brokerage transaction volume without hiring more customer service reps--a strategy that Fidelity hopes will improve its own productivity the same way ATMs help banks cash more checks without adding extra tellers to the payroll.

What's interesting to us about both of these models is that they suggest ways to achieve high growth and sales volume in service automation markets without requiring much investment in building a distribution infrastructure. Moreover, by working closely with large customers, Rosh and MECA have been able to leverage generic product design and marketing skills rather than have to develop grassroots expertise in individual industry segments. Clearly, the barriers to entry in service markets aren't too difficult for PC software companies to overcome. The toughest barrier, we suspect, is probably the perception among developers that electronic cash registers and customer service databases aren't sexy technologies. Granted, service automation isn't especially glamorous--but it is a market that's likely to reward innovation and risk-taking in a big way. For software entrepreneurs, in fact, service automation could be be the single biggest frontier of the coming decade.
COPYRIGHT 1992 Soft-letter
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:automation of service industries will be next major market for software companies; major application areas and models for reaching customers are discussed
Date:May 11, 1992
Previous Article:The 1992 Soft Letter 100.
Next Article:Electronic distribution makes a comeback.

Related Articles
Platforms: a multi-segment market model.
Chase Manhattan selects NCR to automate more than 500 retail branches.
Versata To Deliver Integrated Solution for Rational Rose to Automate Complex Business Processes For B2B Web Sites; Versata Joins Rational's Unified...
Versata Expands System for Automating E-Business Transactions Unique Rules-Based Approach Solves Commerce Change Management Issues.
Versata Launches Interaction Server to Further Automation of E-Business.
QuickArrow Vice President to Speak at ProjectWorld Toronto.
Storage area management: the next generation. (Storage Networking).
Streamlining sales: FreedomNet makes car repair and maintenance easier for dealerships and their customers.
HP to accelerate adaptive enterprise strategy with acquisition of Novadigm and Consera Software.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters