Progress Software History
Progress Software History: A Point in Time (1994/1995)
100% of the credit for the content below goes to Doug Mankoff, Dwayne Rush, and Nick Solomou, students at Harvard Business School who did a paper on Progress Software and Joe Alsop in the 1994/1995 time frame. I OCR'd the content in from an old hardcopy I had and then hand-stiched the appropriate HTML formatting. It's something I've been meaning to do for years. I want to thank Doug, Dwayne, and Nick for their outstanding work.
In addition to the article, you can find more (historical) information about Progress Software here:
Here goes ....
RECOMMENDATION: You may want to print this out as it is very long.
Objective: top management position and equity participation in small computer related company, preferably product and application oriented.
-Resume of Joseph W. Alsop, July, 19811
In December of 1981, Joe Alsop decided to fulfill his objective by starting Progress Software. Eleven years later Progress Software enjoyed sales of $ 111 million, a compounded growth rate of more than 50% per year, and consistent after tax profits of 10% of revenue. Joe Alsop is the CEO of Progress and owns 9% of its stock.
Progress Software Corporation develops, markets, and supports software applications development tools. Its main product, called Progress, is used by applications developers to build "mission critical" software for businesses. The majority of Progress customers are value-added resellers (VARs). Most Progress VARs use Progress to provide computing solutions in the minicomputer or microcomputer environment, as opposed to the mainframe environment. VARs were often consultants who would use Progress to design a unique program. The program might help a manufacturer track orders and inventory or an airport automate its flight plans. Progress makes money from the sale of the program to the VAR and from license fees to the VAR's clients.
While about two-thirds of Progress customers are VARs, one third are MIS departments at businesses and government agencies. To service VARs and MIS departments, Progress employs a direct sales force in North America and accesses independent distributors in thirteen countries abroad. Upgrades and support services account for 30% of Progress revenues. License fees add to this steady flow of cash, resulting in a reduction of risk.
The software business is great. No cost of sales, high growth. The bad news about the market is high fixed costs. If something goes wrong, you are in trouble so fast it makes your head spin.
- Joe Alsop2
The variable cost of a copy of software is minuscule. Extensive research and development on the front end makes the software industry a high fixed cost industry. Typically, the majority of fixed costs come in the form of salaries. Many start-up software companies see no revenues in the first one to three years of its operations. In addition to R&D, software companies must sink significant amounts into marketing- again before seeing any revenue. Applications development software is a small but growing part of the software industry. The users of applications development software, or "development tools" are software programmers. A software programmer's switching costs for development tools is very high. Learning how to use a certain set of development tools usually requires a significant investment in the form of time. In addition, abandoning current development tools requires converting current systems, which requires significant reprogramming.
To a programmer, the relationship with the supplier of development tools is crucial. The programmer needs quick support in the form of a knowledgeable engineer. In addition, the programmer counts on the supplier of development tools to be responsive in the form of frequent and innovative upgrades. When a programmer decides to work with a particular development tool he or she must think of the long term as well as the short term value of the product. If the supplier of the development tool goes out of business or does not have the resources to provide necessary support and upgrades, the programmer is out of luck.
The development tools market is a risky environment. After investing in considerable R&D, a producer of development tools can, and frequently do, arrive to the marketplace to find two horrors. One is that faster competitors have already locked (remember the high switching costs) the programmers. The second is that the marketplace has been drastically altered. Many companies producing software for OS/2 perished when Windows became the operating system of choice. In Joe Alsop's words, "If Bill Gates decides the whole industry takes a hard left, you gotta be ready. [When Gates went with Windows,] Everybody was screwed."
With its low start up costs (virtually no capital equipment requirements), the industry has low barriers to entry. Anyone who wants to program without pay for a couple years can enter. Because of high switching costs, the barriers to entry are more formidable where products already exist. It is to the advantage of new entrants to carve out niches where they can be first movers.
In the minicomputer and microcomputer environments, development tools are produced by companies that primarily offer development tools and companies that offer development tools as supplements to other products. Progress, Powersoft ($52 million in sales), and Cognos ($148 million) are the leading competitors that primarily offer development tools. Recent entrants to the field include companies that made their mark in database management. These are larger companies led by Oracle ($1.5 billion), Sybase ($427 million), and Informix ($353 million).3
Joe Alsop grew up in a family that treasured independence. All but one of Joe's siblings grew up to work for themselves: one is an art dealer, one an author, one a publisher, one an entrepreneur, and one a Bible translator. Joe claims that corporate life was not in his genes. Working in large, bureaucratic organization never seemed an option.
After graduating from MIT in 1967 with a bachelors degree in electrical engineering, Joe went directly to the Sloan School of Business. While at Sloan, Joe and a former MIT professor started Intercomp, a business that produced computer peripheral hardware. Company sales, which varied widely, reached $20 million.4 Whenever the business was going well, Joe would withdraw from Sloan to concentrate his efforts on the business. When the business did not do well, he returned to school full time. Prior to the company's final downturn, Joe went back to the registrar. She, however, decided that enough was enough and denied him re-admittance to Sloan. In 1973 Joe and his partners sold Intercomp to a California based firm. Though a business school drop out, Joe says he "learned a lot of very useful lessons.
Among the lessons was how to manage scarce resources. "With a small company or a small start up, you have to be very good at minimizing your expenses while trying to get your key development work done in the early stages.5" Another lesson was how to deal with venture capitalists. Even though he "burned up a lot of venture capital", Joe insists that he did not burn any bridges. According to Joe, the venture capitalists were OK with the fact that the venture lost money. Joe did what he said he would, and things didn't work out. They knew the risks.
After the sale, Joe went to California to work for the company that bought Intercomp. He left the company a few years later and headed to Dallas, Texas. After a brief stint as an employee of MARC, a market research company, Joe went out on his own as a computer systems consultant. His clients were typically smaller businesses that did not have mainframe computers and operated on other platforms such as mini computers. Joe hired programmers to develop software that met his client's fundamental business needs.
Inception of the Idea
The vast majority of Joe Alsop's consulting clients used non-mainframe computer systems. As a consultant Joe was familiar with the frustrations of software programmers in those platforms. A significant portion of a programmer's time was consumed by repetition. The traditional programming languages (COBOL, C, Basic) required repetitive manual labor. Many program lines were virtually the same for different codes, but the similarity did not significantly reduce the amount of programming required. Similar tasks required entire new sets of code. In Joe's words, the market "heeded a powertool to avoid manual work."
I had this vague notion that programming in the traditional programming languages was a tremendous hindrance to productivity. I had some idea, having done some programming when I was a student, that there had to be a better way. We were all walking, and we needed something called an automobile.
In the world of mainframe computers, others had come to the same conclusion. Several companies had begun to market crude development tools. Joe was certain that similar tools could be developed in the other platforms. Software productivity was a very basic problem, and Joe realized that he could make a lot of money by solving part of the problem.
Team and Company Establishment
Joe was confident that he could build a business around the need he perceived. To be successful, he must bring to market a product before others perceived and filled the need. Joe had saved some money from consulting and from the sale of his previous company. However, it was not enough to hire programmers and a sales force. Thus Joe faced his first start up decision. Should he go directly to financing sources and then hire developers and a sales force? Or should he find people who would donate their efforts in exchange for equity and worry about financing after developing the product? Joe decided to go with the latter option. It was the first of many start-up decisions he would face.
Through his industry connections, Joe met three people who he felt could help him start the company. Two, Mary Szekely and Clyde Kessel worked for a company called Mitrol, where they gained valuable experience creating development tools for the mainframe environment. In 1979, General Electric bought Mitrol6. Szekely and Kessel longed to return to the freedom of a small business environment. When Joe found them, they were open to his proposal. Kessel and the individual who had not been at Mitrol agreed to work without salary for two years in exchange for a piece of the new venture. Szekely become the company's first salaried programmer.
Before a single line of code could be written, Joe and the co-founders had to face a crucial product decision. They had to decide which platform their tool would be developed on. In the computer industry, "platform" refers to the type of computer hardware employed. In the 1970's the most popular platform was the mainframe. For the most part, only large businesses could afford mainframes. By the early 1980's three much more affordable platforms had emerged: mini-computers (called "MINIs"), personal computers ("PCs"), and UNIX workstations.
To Joe there were two crucial criteria in making the decision. The first was that the platform choice had to have the potential to represent a big enough market. Eventually, enough businesses would have to use the platform to support Joe's company. The second criterion was that the platform choice had to represent a market that wasn't too big. Joe's group did not have the resources to compete head-to-head with the larger, established software companies. The ideal situation would be a small but growing niche market. Joe was able to quickly dismiss the mainframe platform. It was the most popular platform among large businesses. In Joe's mind, mainframes represented a mature market. It was big and not likely to grow significantly. While the group had experience writing development tools for mainframes, larger software companies, such as Mitrol, their former company, were already writing development tools for mainframes. The second possibility was the PC. IBM PCs were just hitting the market in 1981. The problem was they were not very powerful. 128k of memory would not be enough to write the kind of software that would help provide important business solutions. The group believed that PCs would eventually get more powerful but not soon enough.
The mini and UNIX workstations were clearly the most attractive choices for Joe's group. In his consulting practice, most of Joe's clients used the mini. It was a platform that many midsize businesses were moving toward. One of the major reasons more companies did not adopt the mini platform was because of risk. To adopt the mini platform was to put all your eggs in one basket. With mainframes and minis, users could only use the operating system that came with that hardware.
UNIX, on the other hand, was a new operating system developed by Bell Labs in relation to an MIT project. This operating system had the unique ability to run on different hardware platforms. To Joe Alsop this separation of hardware and operating system represented the way of the future.
UNIX is not necessarily the technically better operating system. It may be, but that isn't important. What's significant is that UNIX is available for lots of suppliers, so customers have the confidence to buy a UNIX-based solution. If one supplier falls behind in terms of offerings, a person can switch easily to another supplier without a big effort to port applications.7
Because of its newness, few, if any software developers had begun to think about writing development tools for UNIX. To Joe it was the perfect choice.
There was only one problem. In 1981 UNIX was virtually unknown outside academic circles. Virtually no business had even heard of UNIX, much less were using it. So, Joe and his group had to gamble. Would UNIX, because of its inherent advantage, spread to the business world fast enough to support the new company? The founders agreed that it would.
A second major decision that had to be made immediately regarded an implementation language. In 1981, there were two viable programming languages for systems programming. These were Pascal and C (the language that people used to program on mainframes was PL/1). Joe and another founder visited a well regarded engineer at a company called Interleaf. Joe remembers the engineer's advice like a pronouncement from the heavens: "C is the language of choice when it comes to serious system development". He and the other founder looked at each other, said OK? and the choice was made. Little did they know how critical this decision would be to their success.
A Rocky Start
The official founding of the company was in December 1981 under the name of Data Language Corporation. Because the other co-founders were located in the Boston area, Joe and his family relocated from Dallas to Boston. Joe located office space in Billerica. The rent was "dirt cheap" and was to come directly out of the founders' savings. Division of tasks was simple. The three co-founders would develop the software, and Joe would handle business matters. To Joe that meant providing leadership, making sales, developing a plan, and financing expansion.
Before moving into their first office, Joe and the co-founders faced their first crisis. The co-founder who had not been at Mitrol received a job offer from another company for equity plus double his previous salary. To Joe's dismay, but not surprise, he chose to leave. It was a difficult time for Joe. He had just relocated his family (including a newborn third child), and was spending his life savings. Without a critical part of the team, Joe did not feel they could get the product out fast enough. He wondered if he should give up now and return to consulting.
Once again Mitrol provided the person. While at Mitrol, Szekely and Kessel worked with Charles (Chip) Ziering. In addition to working in software development at Mitrol, Ziering also worked in marketing, and had important contacts in Europe. Ziering, who had left Mitrol in 1979, agreed to become the fourth co-founder and to work without salary in exchange for equity. The team was set.
The Non-Revenue Years
The founders moved into their office in February of 1982. Ziering, Kessel, and Szekely immediately began the long process of writing the software. The initial name of the product was SPECTRUM. The goal was to introduce the first version of SPECTRUM prior to the November, 1983 Comdex show. The company planned to make its first shipments in January, 1984.
In the meantime, Joe began his marketing effort. Fortunately, Joe had some experience with the market. It was the same market he served as a consultant in Dallas: small to mid-size businesses that needed tailor-made programs to operate their businesses. The ultimate user was usually the MIS manager of the company.
The conventional wisdom for selling developmental software was to go through the OEM channel. Therefore, Joe began his sales efforts with UNIX manufacturers. To sell the product Joe had to demonstrate it on the company's computer. Because UNIX computers were not cheap, the company owned only one, and it was used by the development team. They frequently had to stop programming so that Joe could use the computer to demonstrated SPECTRUM in some other part of the country. In the end, Joe's efforts proved fruitless. The OEMs already had deals with two other database/too Is companies. These companies provided a product that emphasized the database end of the system, as opposed to developmental tools. Nonetheless, the OEMs were not interested in a deal with a third company.
Without the OEM channel Joe was at a loss.
So I went back and I thought about it. And I thought about why Ashton-Tate had become successful. To learn more, I attended an Ashton-Tate conference, and I had the chance to mingle with a lot of people involved in the startup of the company.8
Ashton-Tate had successfully launched D-Base, a program intended for the PC market. When it found OEM channels closed, Ashton-Tate went after the computer dealers. Their tactic was to offer a 40% discount on the second copy, when two copies were purchased. When people called Ashton-Tate to order D-Base, Ashton-Tate would call a computer dealer in the caller's area asking if they wished to become a D-Base dealer. Ashton-Tate informed the dealer that one sale had already been made, so all the dealer had to do was purchase the second copy at the 40% discount. The strategy worked, and Ashton-Tate had found an effective channel.
The equivalent to computer dealers in the UNIX market were value-added resellers (VARs). VARs were often consultants who added value by using the software to develop solutions for businesses. According to Joe, VARs were previously "beholden to their computer manufacturers and saw UNIX as an opportunity to get out from under the strict dependence on a hardware company.9"
The disadvantage to going with VARs was that Progress would see very little revenue up front. The advantage to going with VARs was long term: Progress would receive license fees as VARs continued to sell vertical market applications10.
Joe learned another lesson from Ashton-Tate. Dealers would buy a product "sight unseen" if the risk was low enough. Companies such as Oracle required dealers to purchase $100,000 of software up front. But such companies had sizable sales forces. Joe decided to make it painless to become a VAR requiring a minimum order of only $3,000.
Because the company did not secure outside financing, Joe had to carefully manage cash. For the first two years, Joe and the other founders did not take a salary. Another way in which the firm's cash was preserved in the early years was by "cutting deals" with both suppliers and customers. The company's "dirt cheap" office space came with a leaky roof and many other shortcomings. Joe used the poor service provided by the company's landlord to save more cash. Whenever some repair was needed, but not provided, Joe would promptly subtract the appropriate amount from the monthly rent. There was little need for "hard assets" (i.e., property, plant and equipment) so capital requirements to actually start Progress were minimal. For the first two years the founders' capital was used to fund operations, which consisted of rent, payments on one UNIX computer, and the travel costs to visit prospective customers.
In February 1984 the company needed to ramp up for introduction of the new product. It raised $1.5MM in exchange for about 40% of the company. Of this amount, $800,000 was provided by two venture capitalists ($400,000 each), with the remainder provided by smaller firms. Joe says getting the funds was "no problem". He simply went back to the venture capitalists that had funded his first venture, Intercomp. Though Intercomp lost money, Joe had done everything he said he would. The venture capitalists fully trusted Joe and were waiting to do business with him again.
RESULTS, OR LACK THEREOF
While product development went somewhat according to plan, the developers did fall behind. To create a product in time for the November, 1983 show, the team had to cut some corners. Data Language Corporation first shipped for revenue in August 1984, eight months behind schedule.
In fiscal year 1985, its first full year of shipments, sales totaled $3.4 million. The company's July, 1983 business plan had predicted sales of $11 million. The market had continued to grow slowly. By 1987 sales grew to only $6.1 million.11 This was one sixth of the amount projected in the 1983 business plan.
The decision to go with UNIX was not looking wise. "The explosion of the UNIX market took a lot more years than we expected12," says Ziering, one of the founders. While its versatility helped UNIX catch on in the business world, it took longer than expected. Prior to first shipments, Joe had counted the number of UNIX boxes in the business sector. A quick calculation showed that he would have to sell software to every single UNIX user in order to break even. As he did not expect anywhere near 100% market share, he could only hope the UNIX market would grow—quickly.
Expansion into Europe
Throughout 1986 Progress Software Corporation (the new name assumed by the company) experienced payment problems with its distributors in Europe. In early 1987, Joe visited several European distributors to see if he could sort things out. The good news was that customers in Europe were extremely pleased with the company's product. The bad news was that four of the key distributors owned by a Swiss holding company were essentially bankrupt. Progress faced a crucial decision. It seemed a shame to eliminate the company's European presence when its goodwill was so high. On the other hand, purchasing the distributors would mean the burden of "more mouths to feed."
A slight rift developed with Board members over the decision. Those opposed to the move argued that the company lacked experience in conducting business in Europe and that its labor laws were too cumbersome. Joe, however, could not get over how positive the European customers were about the product. He convinced the founders to once again "roll the dice." After brief negotiations, Progress acquired the distributors and set them up as four separate subsidiaries of Progress. Progress Software was now a truly international company with offices in Germany, Denmark, Sweden and Norway. The company now employed an additional thirty people. Their mandate was simple: sell enough product to "pay for yourselves". Despite these initial low expectations, the European operations helped Progress hit its long awaited growth spurt. Progress enjoyed a significant first mover advantage in Europe.
At Long Last... Growth!
1988 was what Joe calls the company's "break out year". From 1987 to 1989 sales quadrupled to $25.4 million. Three factors caused the company's overdue success. The most important was the acceptance of UNIX. In the mid 1980's Sun beat Apollo on the workstation game by adopting UNIX early and by standardizing on it. UNIX was no longer a four letter word. By the beginning of 1988, it had found acceptance in the business community. A second reason for rapid growth was recurring revenue. The VAR strategy was beginning to pay off. Royalty payments from VAR customers began to add up. A third reason was the success of the European strategy. Not only did the European distributors manage to feed their own mouths, they began to contribute significantly to corporate profits.
Growth continued throughout the 80s. At one point, because of nervousness about being able to meet payroll one sixth of the company was sold in exchange for $3 million. As it turned out, the company never had to use these funds. The cash produced by sales proved more than sufficient to meet working capital needs.
In 1991, the company sold 1,380,000 shares of stock through an IPO. Of this total, 780.000 shares were sold on behalf of the company. The remainder was sold on behalf of investors owning stock prior to the IPO. From this offering, Progress netted $18.1 million after expenses ($23.25 per share). At the time of the offering, the company was generating sufficient cash to fund both operations and growth (a trend that continues to this day). The primary reason for the offering was to create a market for the company's stock and to facilitate future access to capital markets. As was the case with prior outside financing, these funds were not essential to the operations or growth of the company.
In fiscal year 1993, sales totaled $111.6 million for a compounded annual growth rate of 54.7% over the eight year period beginning in 1985 (see Exhibit 2). Pre-tax profit margins has run 15-17% of sales each year.
The balance sheet of the company (see Exhibit 3) illustrates the conservative financial management of the company. At YE 1993, the company held very little debt ($164,000) on its books. The vast majority of the company's assets are cash and marketable securities. The company maintains this high "cash" balance to guard against an industry downturn. Joe also believes that given the volatile nature of the industry, it would be dangerous to couple high business risk with high financial risk. Furthermore, customers are, to use Joe's word, very "skiddish." They like to know that as a company, you'll be around to service their needs. A conservative balance sheet helps provide that security. Joe continues,
The leverage is in technical innovation and grabbing market share early. That's where the win is in high tech. It's not in financial engineering.
Two of Joe's three co-founders are still with Progress. Chip Ziering is Vice President of Development and Chief Technical Officer. Mary Szekely is the nucleus of the development organization. Clyde Kessel left the company in the summer of 1994.
The future holds a number of organizational challenges for Progress. The company now has forty-five offices and employs over 1000 people. Joe concedes that he "has never managed 1000 people before." He says this poses a different set of challenges.
They include keeping everyone in the much larger organization motivated, developing strategic multi-year plans, and worrying about where you want the company to be down the road.13"
On the development side, Joe believes that the company must continue to innovate in order to motivate and retain talented people. Innovation will also dictate whether Progress can continue to charge premium prices in an increasingly competitive environment. Joe feels that in today's market any product offered by the company must have a combination of "speed and sex." He defines "speed" as quick response to commands and user-friendliness and "sex" as good looks and great graphics on the screen. He continues.
Things are always changing in this industry. You can't be complacent. You must constantly look for what we call the "NBT"-- the next big thing— and ask yourself how you're going to take advantage of it.14"
In the past, Progress has enjoyed success by identifying new opportunities where little competition existed. Unfortunately, the field upon which Progress plays is rapidly filling with companies seeking to exploit the same opportunities. This means that Joe and other top executives at Progress must become adept at managing a company in a competitive environment.
Joe compares predicting where the market is heading in the software business to slalom ski racing.
If you focus on the gate that you're passing, you're dead. Your eyes and your mind must be focused two or three gates down the hill. If you're not thinking ahead, you'll be out of the race.15
EVALUATION OF STRATEGIES
When Joe and the co-founders started Data Language Corporation, they had two things going for them: a good idea and programming expertise. Dan Bricklin had a good idea and programming expertise when he began Visicalc. Data Language Corporation turned into Progress Software, a highly profitable company. Visicalc lost loads of money and was eventually consumed by Lotus, Why the different outcomes? One reason is that Joe Alsop seemed to realize that a good idea and programming expertise did not add up to a substantial competitive advantage. Thus he chose to follow his "small but growing" niche strategy. Had Data Language Corporation gone head-to-head with significant competition early, it might have been crushed. By going after the quiet UNIX market, Joe was able to secure a loyal following and to ramp up slowly.
On the other hand, one might argue that Joe underestimated his company's potential. Perhaps Progress could have developed tools for a more widely used platform and then developed tools for UNIX. This may have required more funding sooner, but securing capital was not a serious challenge for Joe. This strategy would have entailed more risk- something Joe was careful to avoid whenever possible.
Team Strategy with Founders
It was a great alliance for Joe. He got the necessary expertise (and programming) without having to pay much money up front. On the other hand, Joe could have raised money at the beginning, paid the co-founders salaries, and maintained much more control. In addition to the decision to avoid risk, such a strategy may have neglected another success factor. Joe's co-founders put in tremendous efforts to make sure the company was successful. If they did not have the large equity stakes and the feeling of ownership, they may not have pushed the company into the realm of success.
VARs, though a second choice, turned out to be an excellent choice. In the software industry, it is essential to get as large an installed base as possible. Reducing the minimum order to $3000 gave the sales force headaches (a salesperson making a five percent commission invested a lot of effort to make each sale, and earned only $150 on a $3000 sale). Because of license fees and recurring sales, the return from some of these minimum order accounts has been tremendous. In fact, the company still has the same $3000 minimum.
The European decision obviously paid off well. In making the decision Joe did not discount the value of happy customers. To have stranded those customers would have been to cut off years of revenue from a group that was growing due to word-of-mount. Today, Progress enjoys a dominant position in Europe.
In retrospect, Joe feels that he should have brought in venture capital sooner. This would have allowed him the opportunity to go after a second platform- probably PCs- much sooner. On the other hand, the risks would have been greater, and Joe seemed to be looking to minimize risk as much as possible.
Joe says he has no regrets about the financial structure of the company. Zero debt is appropriate because of the high business risk. We are not sure we agree with Joe's extreme position. He underestimated the constant cash flow from licensing and from upgrades. Perhaps some debt would have been appropriate, and, if issued in conjunction with the venture capital, would have afforded Joe more control. Having said this, it is easy for us to look back. Joe had already experienced one failure and he had a family to support.
Advice to entrepreneurs: pick a big fundamental problem and then try to solve a small piece of it so that you can start small and get going. What you want is a $10 million market that is a piece of, and growing into, a billion dollar market, so you can get in now and grow with the market.
Choose an industry niche based on risk profile.
Bill Gates started Microsoft from scratch and now owns 30% of a large company. Joe Alsop started Progress from scratch and now owns a smaller piece (9%) of a much smaller company. Should Joe Alsop harbor some regrets?
He says no. When Bill Gates started Microsoft, he was twenty years old and had no responsibilities. Joe Alsop started Progress in his mid thirties. He had a wife and three children. Joe pursued a lower risk course than Bill Gates. By avoiding the mass market, and the need to throw huge sums at mass marketing, he lowered risk to a level he could live with. At all times Joe avoided competing with the big industry players. Never forget the customer's needs.
Joe was not certain about all his decisions. He was, however, extremely certain about one. He knew that there was a need for development tools in the marketplace. He knew this from years of experience as a consultant. Joe was the kind of person who would later become a Progress VAR. He knew what solutions consultants and MIS managers were looking for.
A significant portion of Progress revenue comes from recurring sales. By always staying close to the customer, Progress had made the necessary adjustments to remain the customer's first choice.
Relationships are crucial in a tightly knit industry.
After coming up with the concept for Progress, Joe thought about who would be the ideal partners. Through industry contacts, he found out about the group from Mitrol. They, of course, proved to be crucial to the success of Progress.
When making the programming language decision, the founders did not rely on their own expertise. They found the leading expert in the software community.
While the advice they received was free, it was invaluable. Over time everyone began supporting C and dropping Pascal for high performance compilers. In Joe's words, "We would have been totally "skrewed" if we had gone with Pascal."
Progress' first customers were people Joe met as a consultant in Dallas. They were MIS managers who had been clients, and they were fellow consultants, such as those he worked with at MARC. Maintaining those contacts proved invaluable. When Joe lacked a marketing vision, he went to Ashton-Tate to find out how they became successful. It was a matter of getting to know the people and hearing their war stories. Again, the advice was free, but proved invaluable.
Finally, despite the failure of his first company he always had good relationships with investors who still believed in him as a person and trusted his professional and technical judgment.
In small clique industries, such as the software industry, it is crucial to be straightforward and to establish a solid reputation. If any of the people who chose to help Joe had heard he was not a "good guy", his job building Progress would have been much more difficult.
Technical Competence Matters when your Customers have Technical Competence
Had Joe not had extensive experience in the computer industry, he would not have been able to build Progress Software. Being able to "talk the talk" was crucial when it came to winning the confidence of the Mitrol group, VARs, and investors.
On the other hand, the founder of Intuit, Scott Cook entered the software business in the early 1980s with no industry experience. He had been at Procter & Gamble, selling Crisco cooking oil. The difference is the customers. Cook's customers had no computer experience. Neither they nor the department store from which they purchased Quicken wanted to know about writing code. Because Joe's clients were programmers he had posses a certain expertise to know how to please them.
Create a situation that accommodates the long-term choices
Joe says one of his few regrets is not raising capital earlier in order to decrease the time it took to bring SPECTRUM to market. On the other hand, not raising capital earlier allowed the founders the freedom to choose long term strategies. Without the pressure to quickly turn a profit, the founders were able to forego short-term solutions such as
Technology platform decisions are unique to certain industries.
Less critical in other businesses, platform decisions are the difference between life and death in the software industry. This is due to the tricky combination of 1) long development lead times and 2) the importance of being quick to the market (due to those huge switching costs). Joe Alsop knew that working without a salary for two years was only possible once for him and the other founders. Because they did not want to compete against big players, Progress had to anticipate trends. The decision to go with UNIX had to be made before they knew whether the business sector would turn toward UNIX.
Decisions must be made when the future is unclear
It seems incomprehensible that the founders of Progress Software chose their implementation language based on the advice of an eccentric engineer who declared, "C is the language of choice.". Had they gone with Pascal instead of C, Joe says they would have been ^skrewed." Does this mean they simply got lucky? Perhaps. But a lesson to take away is that making business decisions, especially in an extremely dynamic market, is a lot different than reading Harvard Business School cases. One does not have the benefit of volumes of research. It helps immensely to have experience, but ultimately one must be prepared to take leaps of faith. Joe Alsop:
At the time you make these decisions, things are never clear. It's a fuzzy combination of experience, research, and gut feel. You have to make the decision when it isn't obvious.
Add to the lack of clarity the pressures of time. In order to get product to the market as soon as possible, decisions had to be made fast. Unfortunately, Progress did not make all the right decisions. There were missed opportunities. Progress did not see the Client Server market until after Powersoft did. Progress did not see the PC market until others did. However, managers in dynamic markets, at one time or another, come to the realization that some opportunities will be missed because some decisions will be wrong. Once a manager realizes this, he or she can resume looking forward toward the next decision.
Financing and Venture Capital Lessons
Joe Alsop does not possess the fear and loathing of venture capitalists that many entrepreneurs experience. This is surprising considering the fact that Joe's first venture was a money-loser for venture capitalists. Joe maintains that venture capitalists will leave entrepreneurs alone as long as they do what they said they would do. Whether Joe was correct or whether he was fortunate to have the right venture capitalists is unclear. If possible, it is a good idea to have two to three venture capital firms. This avoids giving one financial source the power to outvote the entrepreneur. Joe maintains that if you can only find one financial source, your idea is probably not that sound.
Joe Alsop prefers venture capitalists as financial sources. He avoided corporate sponsors because he felt he would have been subject to the corporation's strategy. Should the corporation decide the entrepreneur no longer fits its strategy, it may cut off the entrepreneur. Venture capitalists, on the other hand, will keep investing as long as they feel they will get a good return on the investment.
For Joe, career risk was not a factor. Joe recalls interviewing with IBM when he was an undergraduate at MIT-After 15 minutes Joe determined that working at IBM, or any large organization, was a much bigger risk than starting his own company. There were, however, other risks. Joe's previous venture did not earn any return for his investors and by starting Progress. Joe was risking his ability to raise money for future ventures had Progress not succeeded. Additionally, Joe invested a great deal of personal capital into Progress Software. For two years Joe sacrificed a salary (estimated to have been $150,000 in total) and contributed approximately $60,000 of his own capital.
Today, Joe owns approximately 9% of the company , which has 5,255,680 shares outstanding. The stock, listed on NASDAQ, closed at $37 7/8 on December 8, suggesting the following value for Joe's holdings:
5,255,680 shares X $37,875 per share = $199,058,880 Approximate Cash on Books + $65,000,000 Total value of Progress + $264,058,880
Over the course of thirteen years, this represents a return of approximately 44%, using $210,000 as Joe's investment into the business. The co-founders seem to have come out pretty well also. Immediately following the IPO, the other founders owned approximately 15.2% of the company. The stock was issued at $25 per share which suggests that their stake was valued at:
5,255,680 shares X $25,000 per share = $131,392,000 Approximate Cash on Books + $14,000,000 Total value of Progress + $145,392,000
Over the course often years (leading up to the IPO in 1991), this represents a return of approximately 48%, using Joe's estimate of $450,000 as their investment into the business.
Returns of 44% and 48% seem astronomical at first glance. However, to quantify Joe's or the co-founders' investments in such a manner leaves out many intangibles. Wear-and-tear on the family, the worrying one does at 3:00 in the morning, and the extra effort one puts into one's own business are not reflected in such numbers.
1 See Exhibit one.
2 Interview with Mr. Alsop, November 21, 1994. All further quotations from Mr. Alsop, if not footnoted, are from this interview.
3 Competitive data supplied by Progress Software Corporation.
4 The Culpepper Letter. April 1993, No 110, p. 6.
6 Data Language Corporation business plan, p. 5, July, 1983.
7 The Culpepper Letter. p. 7.
8 New Enterprise Associates newsletter. Interview by Nora Zietz. Data unlisted.
10 Software Magazine, January 1994, Sentry Publishing Co. Inc. Westborough, MA.
11 Progress Software Press Release, June 29, 1990
13 The Culpepper Letter, p. 6.
Email Patrick Lannigan at lannigan at gmail dot com for more information
This page was created and/or refreshed on April 12, 2017 @ 14:50:56