Profile and analysis of a start-up software development company.
Profile and analysis of a start-up software development company.
It's a race against time. Can Howard Smith establish Clarity Software as theworkstation standard before high tech's giants mount their inevitableattack on his growing market?
The beauty of computer software is that it can be developed by just a handful of people. If it's a winner, there's very little overhead to siphon from the profits. But with the computer revolution now in middle age, it isn't easy to come up with a product that hasn't been thought of before. Even tougher is getting the attention of enough users to make a unique program a best-seller. There's simply too much competition.
Occasionally, however, someone does come up with an idea so simple, so elegant, that the market sits up and takes notice. Howard Smith, a tall and animated mathematician cum software engineer, is betting that the product made by his new company, Clarity Software Inc., fits that description.
Two years ago Smith, then 49, was contemplating a well-deserved sabbatical from his employer, Silicon Graphics Inc. (SGI), a Mountain View, Calif., maker of computer boards for three-dimensional graphics. He had come aboard in 1984, when SGI was a $5-million start-up. By 1989 revenues had climbed to a heady $400 million. Smith wanted a rest and had decided to take all of 1990 off.
But no sooner had his wife started collecting travel brochures than Smith became absorbed by a nagging issue at the office. As head of SGI's software products, he oversaw a group that developed complex software for workstations -- expensive and powerful computers that can make several million computations a second and are perfectly suited to engineers. They even use a special operating software, called Unix, that controls the computer's internal instructions. Still, Smith's group was having enormous difficulty communicating with the rest of SGI, which had begun using personal computers.
Deftly outlining the problem on a board at Clarity's modest offices, Smith takes a crimson marker and circles PCs, then Macintoshes, and finally, a large block labeled Unix. "Our group was an island," he recalls. "We'd have to print out copies of what we were working on from the Unix machines and walk them upstairs and reenter them into the other computers. I knew SGI had a problem to which there was no easy solution." And the problem was only going to get worse as SGI continued to grow.
In typical fashion, Smith began tinkering with some potential solutions and finally decided that a new form of software was needed: One that would link Unix workstations to the rest of the computing world. One that could "launch," or facilitate, the use of any popular Unix-based software program such as Lotus 1-2-3 or WordPerfect, convert documents from one application to another if need be, and then send the finished document to another computer, whether it was an IBM machine or an Apple. An engineer preparing a document should also be able to pull into it a spreadsheet, a photo, or even a piece of sound (such as a colleague's comment), and then shoot it off to another department.
Luckily, Smith knew of advanced software programs, developed at a handful of universities, that accomplished some of those tasks. The most sophisticated of them allowed software designers to combine fax, graphics, text, and sound. Better yet, since they were developed at universities, they could be commercialized for small royalty payments. Those programs served as the basis for the type of easy-to-use software Smith would eventually develop.
By November 1989, with his leave approaching, it was clear to Smith that his software idea wasn't something that fit into SGI's product offerings; it had much broader applications. Moreover, there were only two programs available on the market that came close to what Smith was envisioning, and they were quite expensive. With the blessings of SGI's CEO, Ed McCracken, Smith began to talk to outside engineers about what it would take to write his proposed program. "I got absolutely committed, and my sabbatical disappeared," he says. Smith broke it to his wife that, instead of taking a vacation, he might be developing an idea for a new company. A romantic tour of Italy could wait. He took a deep breath and lent his start-up, Clarity, $100,000 from his savings.* * *
Smith was sure he was tapping into an exploding market, and the numbers have borne him out. Dataquest Inc., an industry-research company in San Jose, Calif., figures that workstation sales hit $8.8 billion in 1991, up 19.7% from the year before. That far outstrips the growth rate in PC sales, which suffered a stunning 8% decline over the same period. Meanwhile, there's been a shift in who is buying these powerhouse computers. Traditionally, workstations have been used mostly in the technical market; in other words, they tend to be found on the desktops of engineers and scientists. But according to International Data Corp. (IDC), the biggest surge in demand is in the commercial marketplace: new markets are popping up everywhere, from Wall Street, where brokers use workstations to track stocks, to factories, where workers use them to track inventories. By 1994 IDC expects more than a third of new buyers to be in the commercial world, versus just 8% last year.
With a booming market and few rivals, Smith figures Clarity and its product, Rapport, can quickly become a standard before other companies recognize the opportunity. So far, the two competing software programs -- one from BBN Software Products Corp., in Cambridge, Mass., the other from Applix Co., in Westboro, Mass. -- are proprietary and don't have much flexibility incorporating other programs. Leaning back in his boardroom, Smith indulges in some boasting: "In five years we'll be one of the top 10 software vendors in the industry."
Smith's next move was to put together his story and try to round up the kind of first-round financing he'd need to make Clarity fly. Since he'd never written a business plan, he engaged the help of Thomas Bredt, general partner at Menlo Ventures, who'd been a friend for 20 years, dating back to the time both worked at Hewlett-Packard Co.
Bredt supplied an outline for a plan and a referral to an experienced lawyer, and Smith hunkered down to put his scenario on paper. At night his wife proofread the plan and typed it. Then Smith headed out into the field for the unpleasant task of negotiating for financing. Venture capitalists took notice. Rapport, it seemed, had just the right appeal. "We like products that are easy to understand," says Joe Rizzi, a partner at Matrix Partners, in Menlo Park, Calif., who is also a Clarity director. "And I think Rapport is it."
Another thing VCs like is equity -- as much as they can get. In exchange for a little less than half of his nascent company, Smith gathered $3 million from three funds, including Bredt's, in just 60 days. Smith figured he'd still have room to raise additional cash at a later date, with a smaller portion of the remaining shares -- which, with any luck, would have increased considerably in value. In three years, after a string of profitable quarters, says Bredt, Clarity would be taken public, and it would reap large rewards for the investors.
Now it was just a matter of turning all that planning into a real product. Smith's next move was to persuade SGI to lend his start-up some workstations and provide the references needed to secure a cheap office directly across from Route 101, the main artery that runs through Silicon Valley. Clarity's offices fit the image of the typical high-tech high flier. The surroundings are surprisingly spare -- except for the borrowed $10,000 workstations at most carrels -- and they could benefit from nicer lighting or even the odd plant or two.
But Smith knew that if his start-up was to succeed, decor wasn't where he needed to spend money. In the software business, your biggest capital investment is in building your team. By December 1989 Smith had compiled a short list of executives he might recruit for the company. He recalled that some of the most creative engineers he had worked with in his 25-year career had been Hewlett-Packard cohorts. He began to try to entice them to spend the next several months of their lives writing software codes. Smith's offer was two-pronged: after convincing a prospective executive of the strength of his idea, he promised to match his or her current salary; then he added a small slice of equity to sweeten the deal.
By March 1990 Smith had hired his first employee, a Hewlett-Packard engineer. Then he hired the first of what would grow to a six-person management team. Within six months, Smith says, he had broken the first rule of finance. "I had made the same mistake everyone makes. The $3 million was to take us through to product launch and marketing. It wasn't enough."
Nearly three-quarters of the initial money went to top-tier salaries and equipment, plus royalty payments for software that was used in the development process. The balance was divided between early marketing research, sales plans, and administration fees, including an outside bookkeeper. "This is very typical," observes Menlo's Bredt. "It takes $1 million to $3 million to get the product done and another $3 million to get the marketing done."
Indeed, a series of problems hounded Clarity. For one, Rapport's function is based on pieces of software called converters. They are programs that permit Microsoft's Word, for example, to be translated from its IBM PC version into one that appears in Unix, with the same formatting, spaces, and typefaces. That way users can keep their favorite software and not have to reformat a document every time it runs on a different operating system. All those operations are transparent to the user, but they must work flawlessly, and that means finding the best converters on the market. Though Clarity, unlike some manufacturers, doesn't rely on a broad array of vendors to make its product, it still contends with the issue of quality. Now Smith suddenly had to deal with being a small, unknown start-up with little clout when it came to asking his suppliers to solve a problem.
"We can turn a problem inside the company around in a week," Smith grouses. "But external problems can take months. You quickly find out that you're number 15 to be called back. It took 25% longer to get a product out."
Because of the delays, Smith and his chief technical officer, John Busch, didn't have a working version of Rapport ready until June 1991. At that point, funds in Clarity's coffers were running perilously low. Luckily, what Clarity had developed proved practically eye-popping to the venture community. Rapport incorporated all the functions Smith had wanted behind a series of pull-down menus, and it painted stunning presentation graphics in 128 colors.
The original backers signed on for a second round but wanted to spread the risk, and suggested finding a fourth fund to share in another $3.5-million investment. When word got out, more than 20 companies asked for a demonstration and a copy of the business plan. Again Smith had to chip off some of the company's equity, but a much smaller fraction than before, since Clarity's value had jumped based on the prototype.
Besides a dazzling software display, the potential second-round investors saw a revised -- and much more modest -- business plan. The first copies of Rapport were shipped in July but got off to a slow start, in part owing to the lingering recession. Smith realized his early sales weren't going to come near his original projections. At the start he and Gwen Peterson, vice-president of marketing and a crack researcher formerly with Dataquest, had estimated that first-year sales would be $10 million. But the economy wouldn't cooperate. Not only were budgets being cut; there was no guarantee that the decision maker to whom you were selling today wouldn't be on the unemployment line tomorrow, and the sales pitch would have to begin all over again. Now Smith soberly projects sales for the fiscal year ending June 30 to be $3 million.
About this time, the clarity management team began to realize that its original marketing plan was also askew. When Rapport was first conceived, everyone figured it would be sold to one person in a department. It was priced just under $900 in a conscious effort to keep it below the $1,000 cost of a typical Unix program. The thinking was, once all the people around saw how easy it was to use, they'd clamor for copies of their own. "Wrong," says Peterson. "Almost all of our sales are to whole departments or to an entire company." That means a longer sales cycle, different manuals for group sales, and, more important, different pricing. Now Rapport buyers typically get a 50% discount for orders of more than 50 copies. That revised strategy shows up in Clarity's projections, in which the average per-unit price decreases each year.
Since Rapport -- like all Unix software -- isn't available in any retail channel, Clarity's seen little reason to do much advertising. Rather, Smith is shelling out his promotional dollars to create a presence on the trade-show circuit. At last year's Uniforum, a trade show for Unix users, Clarity spent $60,000 for a booth and attracted 600 sales leads.
Much of the word about Clarity goes out through the sales effort, which engages one-third of the company's 32 employees. Each of the six salespeople shepherds one channel of distribution, from the value-added resellers, who include Rapport in large contracts, to the hardware makers themselves. Smith has fielded two additional salespeople in Europe, who work with 12 distributors there. Three people work in telemarketing, shuttling queries and leads to distributors in their areas. Though Rapport is available only in English today, Smith plans to have French and German versions by this summer.
But Clarity is garnering the most success by riding the coattails of workstation sales and vigorously pursuing distribution through original equipment manufacturers. A trend in Clarity's favor is that large corporations are switching from giant mainframe computers to a string of workstations -- in short, swapping a centralized and expensive piece of hardware for a number of cheaper, and in some cases faster, machines. That's forcing workstations into areas that normally would accommodate PCs linked to a mainframe. Also, some businesses, like brokerage houses, are souping up their computing power to run ever more complex programs. They too are turning to workstation vendors like Sun Microsystems Inc., Digital Equipment Corp., and Hewlett-Packard. Increasingly, offices in Fortune 1,000 companies are becoming mix-and-match collections of PCs, Macs, and Unix computers.
Paul Schumann, Clarity's vice-president of sales and the former head of BusinessLand's U.S. network, arranged for the top workstation makers to refer leads to his group. For instance, Sun -- the industry leader -- is making its biggest gains selling outside engineering departments, which had been its traditional niche. Rapport, with its ready ability to share documents with PCs and Macs, is an attractive package for Sun to push in its own sales pitch. And this past January, Clarity finalized a deal with San Jose based Acer/Altos to have Rapport bundled for sale with the hardware maker's workstation systems.
To make it easy to try Rapport, Schumann offers the program free for 30 days. If the user likes it, he or she can buy a license through an electronic-mail number, and with it an entry code. Otherwise, after a month the program can no longer be entered without the code and becomes useless. Once an order is placed, Clarity promises to have copies out the door within 24 hours. Since software is digital data stored on magnetic tapes, duplicates can be easily made and there's little reason to keep large quantities in inventory. Clarity, like most software companies, uses an outside supplier who stands by the phone when a big order is pending. Overall, Clarity has shipped around 4,000 units to date.
Still, sales are falling well short of Smith's initial goals. And since his volume discount and OEM distribution strategies will require selling relatively greater numbers of units with each successive year, Smith's biggest challenge is keeping a tight rein on expenses. Despite lower revenues, he intends to be in the black for the fourth quarter of 1992, as planned. Jim Lawrence, Clarity's chief financial officer, is trying to make sure that expenses don't creep up any faster than half the rate of sales. All the top officers have forgone raises, and for the first year, Lawrence even used an outside bookkeeper. Tight times means there was no Christmas party, and hiring was frozen back in July.
Smith admits he's stretching people as far as he can. For technical support, critical in helping customers familiarize themselves with a new product, there is an electronic-mail service for customers to use if they have questions. If it's a routine query -- as most are -- a prepared message is sent back to the customer. "Our support team is 2 people," says Smith with a laugh. "Folks think we have 30. The edge is our automated support by E-mail." Also, he admits, those two employees work 12-hour days. Fortunately, there's little aftermarket service needed for Rapport. Once customers get a copy, they start customizing the program to their own needs.
All those measures keep a tourniquet on spending, and Smith doubts he'll have to go back for additional financing. In an industry where 30% margins are not unheard of, Clarity's are a respectable 12% to 13%, and the company is able to plow a third of sales into research and development.
Smith had better not scrimp there. He must address complaints leveled against the early version of Rapport. To buyers used to PC software, Rapport's spreadsheet and word processor seem inadequate. For example, writers miss the ability to footnote, and there's no thesaurus. For engineers and researchers using Unix, the communications features are nice, but not nice enough to make Rapport an easy sale. "What we need is a little more maturity in the product," admits Clarity's Washington, D.C., sales rep, Blix Jones. "Customers are saying, 'I like what you have, but I'd like to see this or that,' and they hold off." As a result, several features will be added this year.
Smith describes the present version of Rapport as a backbone for a number of different tools. The latest is a program that lets the user both send and receive faxes. Next he hints at adding video and other features to meet the burgeoning demand for multimedia communication. The icing on the software would be a communications program that lets users call up an electronic profile of the intended recipient. It would list that recipient's electronic-mail address, fax number, and even the type of word-processing program he or she uses, so the final document could be sent in an easily read format.
But no amount of R&D will address Smith's biggest concern. Computer analysts point out that it's only a matter of time before a large, well-established software company, like Lotus or Microsoft, comes into Rapport's market. Though both sell almost exclusively in the IBM and Apple arena, the leaps in workstation sales may be enough to lure them onto a collision course with Clarity. What Smith is banking on is a two- to three-year head start against those deep-pocketed rivals. The aim, he argues, is to establish Rapport as a well-entrenched standard. By 1994 he hopes sales will hit $20 million, providing for a robust R&D program. At that point, he should be able to take Clarity public and boost his presence even further with the proceeds. "Even solid entrepreneurial companies run scared," says Smith with a smile. "You always have to be looking over your shoulder."* * *
Matt Rothman is a free-lance writer based in Los Angeles.* * *
Clarity Software Inc., Mountain View, Calif.
THE CONCEPT: Develop easy-to-use Unix-based software programs such as a word processor and a spreadsheet, and combine them with a communications package that links workstations to IBM and Apple personal computers; price the product well below typical Unix programs in order to establish it as a standard
PROJECTIONS: A loss in 1992 of $2.2 million on sales of $3 million; 1994 earnings of $2.4 million on sales of $20 million
HURDLES: Building a large enough customer base to become the first choice in communications and productivity tools for Unix workstations, before bigger software companies spot the opportunity; continuing to introduce new features and enhance existing ones
Family: Married, five children
Source of idea: At Silicon Graphics Inc., he saw a growing presence for workstations in the workplace and a growing need for them to be more versatile and better integrated in that environment
Personal funds invested: $100,000
Equity held: A "substantial" piece of the 50% equity that is owned by Clarity's employees
Education: Attended Los Angeles's City College; B.S. and M.S. in math from California State University in Los Angeles and San Jose
Other companies started: None
Last job held: Vice-president and general manager of Silicon Graphics' work-group-products division
Clarity Software Inc.'s Projected Operating Statement ($ in thousands)
|Fiscal year (ends June 30)||1992||1993||1994|
|Average price per unit||$.750||$.682||$.555|
|Cost of sales||$300||$1,500||$3,000|
|Research and development||$2,800||$6,100||$7,200|
|Sales and marketing||$1,100||$4,100||$5,100|
|General and administrative||$1,000||$1,800||$2,300|
|Pretax Profit (Loss)||($2,200)||$1,500||$2,400|
WHAT THE EXPERTS SAY
Editor of P.C. Letter and editor-in-chief of InfoWorld , a weekly news-magazine for PC-systems managers
There are two ingredients to success for a high-tech start-up: an understanding of the customer's problems and needs, and an understanding of the technology. Many new companies -- and Clarity is definitely in this category -- focus too much on the latter and don't pay enough attention to the former. That is, they're not customer driven. In a way, it was unfortunate that Smith was so well funded, because it enabled him to throw a good deal of his money into engineering. Had he been forced to bootstrap a bit more, he probably would be more customer driven now.
Smith's background is with hardware companies, so he took a hardware approach to building software. Essentially, Rapport was created by pasting together a bunch of different applications. And the product's value is in that collectiveness. The problem is that, taken singly, none of those applications is as good as other stuff on the market.
Smith had two choices for his marketing strategy: sell Rapport as a productivity program that integrates a bunch of disparate applications for Unix, or sell it as a communications program -- the first real E-mail package for Unix. In my view, the clear value in Clarity's product is as an E-mail program. But Smith chose the productivity angle. That was a mistake because by the time Clarity came along, there were already several products on the market that offered that. That marketing miscue, and not the recession, is the reason for Clarity's low initial sales.
There's still time for Clarity to reposition its product. The window of opportunity is not as small as Smith might believe. Clarity's market is real, but it's too small a niche to interest competitors like Microsoft or Lotus, which are used to selling hundreds of thousands of units per month. So Clarity's niche is there, but it must learn to listen to its customers. If it does that, nothing will be able to put it out of business.
A partner at Institutional Venture Partners, a $400-million investment fund. Yang focuses on software and communication-hardware ventures
Overall, Smith has a good idea. This is definitely an emerging market. Typically, people bought Unix machines for a single application -- not for overall productivity -- in science environments. Smith is coming in and saying people need group-ware -- productivity tools for multiple users. But there's a question of whether people will buy a software product that horizontally integrates several applications, none of which is the best quality available on the market. Second, Rapport is trying to communicate, and coming up with the converters for that is no small task.
The big issue here is sales and marketing. From a technical point of view, you understand why someone would use the product, but figuring out how to sell it is difficult. There are no established distribution channels. There are no retailers and few value-added resellers for Unix.
Selling through Sun and other vendors makes sense, but those guys are really more interested in selling hardware. Besides, they themselves still don't quite know how to crack the code and get into the commercial market. They probably won't have the knack for another two years. With group-ware, it's tough to find the decision maker and to get a consensus among all users at a company.
You could blame the slowness of early sales on the recession to some extent, but I'd argue that it has had more to do with the one-two punch of trying to pioneer new applications and new distribution channels at the same time. Smith's projections look reasonable, but the sales growth called for is especially aggressive, considering the fact that the average price per unit goes down each year.
This is going to be a make-or-break year for Clarity. It will be too early for it to crank up expenses so it can do further product releases. I don't think it'll have the money to continue to upgrade the product in the short run. I'd advise driving to break-even until it has solved the distribution issue. It would be foolish for Clarity to increase engineering and advertising costs until it has done so.
Finally, the emerging competitive landscape will be a big worry for Smith. If Unix workstations really begin to take hold in the commercial market, Clarity will have a host of competitors. But on the other hand, Microsoft is betting against Unix's taking over, and it's working on technology that will offer Unix capabilities through DOS interfaces. Either way, it'll be tough.
President of Applix Inc., a seven-year-old company that sells software applications for workstations
There's definitely an opportunity here. But to succeed, you first and foremost have to introduce a great product that adds value. The end-user has very high expectations. Nobody's buying workstations to get a better word-processing program or a better spreadsheet. Workstation software has to add new tools. I don't see that with Clarity's Rapport. The company will need more time to develop more sophisticated capabilities, and that may mean it will miss its window of opportunity.
Right now I don't think Clarity's numbers are great. There really hasn't been a recession in this market to justify them. Its pricing strategy is similar to ours, and I think it's the right one for the market. But the test will be over the next 18 to 24 months. Again, the key is the product -- if it can be upgraded and is being received well by customers, Clarity has a chance at hitting its 1994 projections.