Software: The Growing Gets Rough

Special Report

One-Product Companies Won't Survive the Fight for Corporate Clients

By Anne R. Field and Catherine L. Harris in New York
With Barbara Buell in Boston, Richard Brandt in San Francisco, Scott Ticer in Los Angeles, and bureau reports
Business Week

March 24, 1986

It's summer, 1978. Most people see the world's first personal computers for consumers, just introduced by Apple and Tandy, as high-tech toys -- and they're selling slowly. Mitchell D. Kapor, who will one day become president of the biggest name in personal-computer software, Lotus Development Corp., recently stopped teaching transcendental meditation. But a few farsighted young software developers already perceive great potential for the personal computer -- which IBM won't add to its line for another three years. One is Daniel S. Bricklin, a bearded graduate student with one of the first Apple IIs at the Harvard business school. While riding a bike on Martha's Vineyard, Bricklin reaches an important decision: He'll go into business for himself selling an imaginative electronic-spreadsheet that he dreamed up at school. A year later, when his program is introduced commercially as VisiCalc, it will make a bigger splash than he ever thought possible. For the first time, VisiCalc will turn personal computers into invaluable tools for the average business person.

In the seven-and-a-half years since Bricklin presided over the birth of the personal computer software business, it has blossomed into a $5 billion industry with 14,000 companies and 27,000 different products. The questions now: How long will it take this industry to shake out and grow up? And how many companies can survive the next challenge -- cracking the big corporate accounts that are becoming the major market for personal computer software?

The answers: not long and not many. Companies that can't keep pace already are dropping by the wayside. Last year, 57 personal-computer software companies were bought out, up from 23 the year before, according to Broadview Associates, a market research firm. Many others died quietly.

The early informal tenor of the software business, set by a group of intense, creative, and often whimsical young entrepreneurs, is quickly giving way to professional seriousness. Software companies realize that they now have to play ball with corporate clients -- who buy about a third of their products -- instead of with a bunch of high-tech enthusiasts. No longer a "rebellious teenager," the personal-computer software industry "is entering a stage of maturity," says Edward M. Esber, president of Ashton-Tate, an industry leader.

This new world is full of big challenges. Until recently, software companies made their money selling to individuals who used personal computers mostly on the job-- and had the discretion to buy the programs they wanted. Now corporations want to tie their computers into information networks. Data processing managers have taken over the lion's share of software purchases to ensure all the machines work together.

These managers are a hard sell. Before they O.K. a program, they want to make sure the company that wrote it will survive to upgrade its products and provide service and support. This is forcing software makers to develop better quality control, special corporate sales forces, and high-volume pricing policies. "The package has to be from someone reputable; it's got to be bug-free; the manual's got to be good; and it's got to be teachable," says G. M. K. Hughes, vice-president in charge of systems and communications at Pfizer Pharmaceuticals. "And it's got to have some reason to displace what we already have."

ONE-PRODUCT COMPANIES

There's just one hitch. In the quirky software business, developing the stability, breadth, and financial staying power that corporations want isn't easy. "There are a lot of companies around that are technology-driven. They're formed to exploit one specific idea," notes John D. Page, vice-president for corporate research and development at Software Publishing Corp. "They're not formed as an engine to keep having ideas." Such one-product companies have often found it impossible to keep the creative juices flowing, to think up and produce follow up products that will let them survive, and grow big enough to win corporate acceptance.

Slower industry growth isn't making things any easier. Over the next five years, market researcher InfoCorp expects personal-computer software sales to grow at only a 15% compound annual rate, down from 64% yearly growth over the past five years. Most new companies are being forced to go after small niche markets or to try to sell their products to one of the bigger players (page 139).

This is leaving the bulk of the market to a few winners. Last year the top 15 companies in personal-computer software accounted for 72% of sales of all general application programs, up from only 37% in 1981. The products of the three biggest companies -- Lotus, Ashton-Tate, and Microsoft -- accounted for 35% of the revenues.

Before long, predicts William H. Gruber, president of Research & Planning Inc., a computer consulting firm based in Cambridge, Mass., the industry will likely split into two tiers. Several major competitors will sell "the central utility" -- all the basic software needed for word processing, spreadsheets, graphics, communications, and data-base management. A second tier of small companies will write programs for specific jobs in niche markets too narrow to concern the larger companies. And they will "piggyback" their software onto the central utility, using similar commands so that people don't have to learn new ones for each program. Providing coordinated programs will be essential, Gruber says, because it is too time-consuming and expensive for corporations to train employees to use software from many different companies. Learning to use Lotus 1-2-3, for example, can take 20 to 40 hours. "There's an awesome cost to training and retraining 100 or 500 people," Gruber notes.

Already, Microsoft Corp. is trying to solve this problem. Its $95 Windows program acts as a layer between the operating system -- the software that guides the basic functions of a computer -- and applications programs for jobs such as word processing. This shields the computer's use from the complex commands usually required to run software. To issue a command, the computer operator has only to point to symbols on the screen, using a hand-held "mouse." Windows, which was inspired by the software run on Apple's easy-to-use Macintosh, also allows the display of several different software programs in boxes, or windows, on the screen, making it easy to exchange data between programs.

Developing such a product isn't easy. Both International Business Machines Corp. and Digital Research Inc. created similar programs -- Topview and GEM, respectively -- which weren't well received.

Software makers that can't reach corporations may still be able to survive by selling through retailers to individuals, who now make up about half the market for personal-computer software. But they may have an increasingly hard battle in retail channels. The players that succeed in selling to corporations will be the ones with money to throw at advertising and other consumer inducements, such as telephone "hotlines" for solving problems.

SEAT-OF-THE-PANTS

This is already making retailers skeptical of programs from small companies whose products generate little traffic. "It's more profitable for them to push fewer and fewer products," says Peter Gabel, president of Arity Corp., a small Concord (Mass.) software company. Concludes William L. Coggshall, president of market researcher Market Access International: "The rich will get richer."

If that's true, the Big Three -- Lotus, Ashton-Tate, and Microsoft -- stand to thrive. All have enough resources -- Lotus has $90 million in cash -- to generate stable revenues and plan ahead for new products. And all are searching for ways to diversify and broaden their revenue bases. Each has already moved beyond seat-of-the-pants management to more professional operating procedures.

Still, these companies have to overcome many of the same problems that bedevil their small competitors. For example, they depend heavily on the creativity and imagination of software developers who don't live by a timetable. That makes scheduling the new product releases -- and planning -- very difficult.

These companies also depend too heavily on a few strong products. At Lotus, 1-2-3 brings in more than 60% of sales -- and the market for spreadsheets is expected to slow drastically. Ashton-Tate's dBase data-base management products contribute 64% of the company's revenues. Microsoft's operating systems bring in about 50% of its sales.

In addition, all three must deal with new distribution and pricing problems. Corporations are more interested in buying directly from them instead of through retail channels. That means the software makers have to create expensive direct-sales forces and increase the number of sales and support people who understand the complex requirements of corporate computer systems. On pricing, corporate data processing managers, who are used to making volume discount deals with mainframe software companies, are demanding similar agreements from personal-computer software makers. They are balking at spending around $495 each for, say, 100 copies. Many are demanding so-called site-licensing agreements that would allow them to make as many copies as they need. So far, the Big Three have refused. "Site licensing is economic suicide," contends Ashton-Tate's Esber.

To meet these challenges and attain the heft they need in the corporate world, each of the major companies has devised its own strategy.

* Lotus is diversifying by buying companies that sell new services to its customers, who have bought 1.4 million copies of 1-2-3. Lotus' eight-month-old Lotus Information Network Corp. unit sells stock market data via special radio receivers and channels it directly into a 1-2-3 spreadsheet. This month, Lotus bought part of InfoCenter Software Inc., getting software that could help personal computers pull data from mainframes into 1-2-3 spreadsheets. More such acquisitions are likely. "It's a gigantic opportunity," says President Jim P. Manzi. "And the entrenched [software] company has all the advantage."

Lotus also has set out to protect itself from the vagaries of the software business by setting up Lotus Magazine to reap revenues from advertisers eager to reach its large customer base. And it has established a division to create special products for scientific and engineering customers, big buyers of its generic products. In addition, Lotus has been reaching for wider international distribution, funding promising software ventures, and financing long-range research in advanced technologies (box, page 133).

* Ashton-Tate wants to position itself as a one-stop shopping center for corporate applications software. The Torrance (Calif.) company bought MultiMate International Corp., which sells a popular word processing program, for $19 million last December. And, like Lotus, Ashton-Tate has started to explore the market for information services, including possible arrangements with such information suppliers as Dun & Bradstreet, Dow Jones, and McGraw-Hill.

President Esber's overriding goal is to turn Ashton-Tate into a computer-services company that can help build corporation-wide systems. For instance, he is bent on developing more sophisticated software that can link personal computers with mainframes so that information can be easily passed back and forth. "The trophy now is being the leading computer-services and products company, not the best micro, mini, or mainframe [software] company," he says. To make this transition, Esber plans to build a staff of 20 engineers experienced in developing major computer systems.

* Microsoft derives a steady stream of revenues by supplying the operating system for all IBM PCs. But it, too, is diversifying. Chairman William H. Gates has started a new division to develop software on compact disks, the silvery platters used for audio recordings that recently have been adapted to run with computers. With Windows, he is aggressively expanding Microsoft's presence in the market for application programs. Gates hopes that Windows will make Microsoft the dominant company in the field and become a new industry standard that other programmers will write for. That way, Microsoft will get license fees from developers designing applications that take advantage of Windows' features.

Although all these strategies are bold, there are no guarantees for any of them. For example, Esber's goal of developing software that links personal computers to mainframes will be hard to attain. "Personal-computer software companies have worked in a very limited environment," says Martin A. Goetz, president of Applied Data Research Inc., a mainframe software supplier. "Mainframe software is a whole different world."

Furthermore, the information-services business that both Lotus and Ashton-Tate are eying is highly competitive. Much larger companies, including IBM, American Telephone & Telegraph, Citibank, and Merrill Lynch, are jockeying for position. "This is one of the areas where the pioneers might end up with the arrows in their backs," concedes Ashton-Tate's Esber. Adds Paul Cubbage, an analyst with market researcher Dataquest Inc.: "It's definitely a lottery ticket for them."

MAINFRAME CHALLENGE

Despite their lead in the corporate sweepstakes, the Big Three cannot count on being the only ones wooing more sophisticated customers, either. A group of second-tier players -- including Software Publishing, Microrim, and MicroPro International -- also are racing to keep up with the industry's rapidly changing dynamics. Software Publishing grew steadily into a $37.2 million company in 1985 by selling its PFS series of seven easy-to-use business programs. But many of its customers have grown out of PFS. Having watched first-quarter earnings drop by 76%, to $617,000, on a 45% decline in revenues, to $6.9 million, it's now pinning its hopes on a more sophisticated line of management-oriented programs. They will be marketed under the label of Harvard Software Inc., which Software Publishing bought for $4.2 million last July. It also will split its PFS products into several levels, one for first-time computer buyers, and others for customers who want more powerful software to work on networked computers.

Microrim Inc., a $15 million company in Bellevue, Wash., sells a $700 database management program called R:base. It has increased its direct-sales force from 8 to 23 people and established a $5,000-a-year corporate service plan. The plan provides corporations with special service and support, product updates, three days of training, and an 800 number for customer questions.

Another challenge for personal-computer software developers may come from mainframe software companies trying to move in on their newly won turf. Mainframe companies have been slow to do that because they don't have experience with customers outside the data processing business. But now they're starting to move. Applied Data Research is developing a version of one of its data-base management programs for personal computers. There are a growing number of alliances between personal-computer and mainframe software companies aimed at linking different types of computers. Management Science America Inc., an Atlanta mainframe software company, has an agreement with Lotus to include its products as part of MSA's mainframe-to-micro software package.

Even though the battle for corporate hearts and minds has become the major campaign in the software industry, a few iconoclasts are proving that they can fight and win a different war. Privately held Borland International Inc. claims to have grown to $30 million in revenues since 1983 by selling inexpensive programs -- usually for less than $100 -- largely by mail. Another company, two-year-old Dac Software Inc. in Dallas, uses the same approach to sell low-priced accounting programs. The $8.5 million company's sales are growing by 10% to 15% a month.

Eventually, if companies like these can hold out long enough, they could end up back on the front lines. If, as consultant Gruber suggests, a few major "utility" companies do succeed in establishing umbrella software standards, some small independents may be able to use those standards to write knockoff programs. They could win respectable shares of the market by underselling the majors. There is already a precedent for such a scenario: cheap IBM-compatible PC "clones" that are winning an increasingly large following today. If a similar software market develops, the dynamics of the business could take a new turn -- toward price wars and a resurgence of small software shops.

BLURRING DISTINCTIONS

Faster, more powerful computers may also enable new software hopefuls to develop programs that perform entirely new functions. It was, after all, the increased memory capacity of the IBM PC that allowed Lotus' Kapor, who is now chairman and CEO, to develop 1-2-3, which combines spreadsheet, word processing, and data-base management in one package. "There will be continuing opportunities for a small number of new firms that could potentially become very large [companies]," he says.

Of course, the personal-computer software industry is so young that it's impossible to discern exactly what it will look like a decade from now. Ultimately, the distinctions may blur between personal-computer and mainframe software companies as they start to enter each others' turf. Probably a few leading players will continue to dominate. There will probably also be a group of small, specialized outfits that dovetail their products much more closely with those of the majors. But for the many personal-computer software companies that aren't able to adjust to today's increasingly complex corporate market, growing up won't be just hard to do -- it will be impossible.

GRAPHIC: Illustration, no caption, ANDREA BARUFFI; Graphs 1 and 2, AS GROWTH SLOWS IN PERSONAL-COMPUTER SOFTWARE . . . THE MARKET LEADERS HAVE TAKEN A BIGGER AND BIGGER SHARE, DATA: INFOCORP, DERRICK A. LANGSHAW/BW