by W. E. Pete Peterson
Copyright 1993, 1998 W. E. Peterson
Going to War
We started 1987 with a 30% share of the word processing market, well ahead of Micropro at 16%, IBM at 13%, and Microsoft at 11%. That share represented a tremendous amount of money, enough that we were finally beginning to understand how important WordPerfect was to our future. Instead of trying our hardest to shed our one-product image, we saw a much greater potential if we would concentrate on word processing. We were not totally abandoning our other products, but having made it to the top of the word processing heap, we were going to make sure we stayed there before we went after other markets.
Our new marketing strategy had three parts: first we wanted to write a version of WordPerfect for every platform that made sense, then we wanted to make sure WordPerfect was well integrated with the other important products on each platform, and only then did we want to attempt to create other products for each environment. On the PC this meant making sure WordPerfect worked well with Lotus 1-2-3 before we worried about making MathPlan a success. On the VAX this meant having WordPerfect work well with All-in-One before we worried about writing our own electronic mail for the VAX. By focusing our efforts on WordPerfect, we felt we could provide a word processor which would satisfy almost every customer on almost every machine.
This idea was out of step with the image that a lot of writers for the trade publications seemed to be promoting. They were writing what I called "Mr. Rogers reviews." Like the song from the Mr. Rogers television show for children which said "everyone is fancy, everyone is fine," the reviewers seemed to feel that every word processor had something valuable to offer.
One PC magazine liked to divide word processors into categories for personal, corporate, and professional use, and publish an article on the best products for each category. The personal category was for housewives and high-level executives--those people who did not need a lot of powerful features and who generally had poor typing skills and too short of an attention span to learn how to use the computer properly. A product like PFS:Write was cheap and more than sufficient for the computer-challenged. The secretary was a typical user in the corporate category. She was described as a "heads-down" user, or someone who spent the entire day sitting at the computer with her head down, looking at the keyboard, typing away like crazy. The standard corporate user needed more features than the personal user, but she did not need all the features of the professional writer. Since she already knew how to use a Wang word processor, Multimate was considered a good enough choice. The professional category was for elite computer users, those with special needs who were knowledgeable enough to use the powerful and complex products like Microsoft Word, WordPerfect, or Xywrite.
This type of analysis made me angry. More than a few housepersons and executives knew how to type and use computers, and I thought a secretary deserved to use the best product available. I felt that anyone who was serious about their writing should be able to use a professional-level product.
I believed a business was much better off finding one word processor that would satisfy everyone rather than a number of word processors to meet individual needs. Documents were not just printed and sent off in the mail. They were passed around, revised, reworked, and published as computer files without ever appearing on paper. To do this easily, an organization needed to have compatible documents in all its departments. Even if an executive never added a footnote to a document, he still needed to read a footnote if the legal department sent him a brief containing one. The legal department may not have wanted to write equations, but if an engineer in another department included one in a patent proposal, then a lawyer's word processor would have to support them.
In addition to having the same word processor throughout an organization, it helped if the same word processor was used all over the place. If one law firm worked with another on a case, their documents needed to be compatible. If a businesswoman was away from the office, she wanted to find the word processor she used at work on her home computer, her laptop, and at her hotel's business center. If a business needed temporary help, it needed its agency to send someone who already knew how to use its word processor. The world did not need the incompatibilities which came with many different products. The world needed a standard.
WordPerfect was in the best position to become the world's word processing standard. We were practically on every machine, while Microsoft was only on two of the little ones and IBM was only on IBM machines. If we could get 5.0 right, we had the potential to stay on top for many years to come.
Getting WordPerfect 5.0 right was not an easy thing to do, however. Wang had proven a company could lose the number one spot by moving too slowly. Micropro had proven a company could lose it by moving too quickly. Somehow we had to find a way for 5.0 to support the new technology without making our old customers unhappy. We had to make sure that the improvements in our new product were attractive enough to overcome the resistance our old customers had towards installing new releases. If 5.0 was too different, too slow, or required too big of an investment in new hardware or new training, we could get into trouble. Since we knew when we started that WordPerfect would have to be a little different, a little bigger, and a little more complex, we had to be careful.
The initial focus of 5.0 was to improve WordPerfect's printing. The new printing capabilities were designed not only to take advantage of the laser printer, but also to take advantage of any special features built into almost any printer. Our plans for the release were expanded, however, when Dave Moon came to the Board shortly after starting his desktop publishing project. He had come to the conclusion that most of what he and his group were planning to do could be included in WordPerfect 5.0. Instead of working on a separate DTP product, Dave wanted his group to add support for graphics into WordPerfect 5.0. Dave believed his group could give WordPerfect users the capability of doing 75% of what a desktop publishing package could do.
His proposal meant giving up on the desktop publishing market, but if we followed his advice, we would jump ahead of what we heard Microsoft was planning for their next version of Word. We would also be able to offer software which had an obvious advantage over anything available from the dedicated word processing vendors. Giving up the DTP market to improve WordPerfect was not a big sacrifice, considering how much trouble we had selling our other products. Dave's suggestion fit right in with our new word processing strategy, so we told him to go ahead with his plan.
Adding graphics support was no small effort. Integrating text and graphics had been an unrealized dream of word processing companies for years. To do it right, we had to wrap text around the graphic boxes, and allow the boxes to sometimes stay in the same place on a page and sometimes move with the text. The graphics could take up a lot of space on the disk, so we had to find ways to accommodate the large files. Since DOS had little support for graphic images, we had to build all the system software pieces to display and edit the figures.
The people in the marketing department went crazy when they heard graphics support was coming. It would be a lot of fun selling a product which was way ahead of the competition.
There were thousands of design decisions which had to be made for 5.0. One of the toughest was deciding whether or not to try to make 5.0 a full WYSIWYG product. WYSIWYG, pronounced wizeewig, stands for "what you see is what you get." This was the ideal situation in word processing, to have the screen, "what you see," look exactly like the printed page, "what you get." When we started in the wp business, WYSIWYG meant showing bolded and underlined text, line endings, and page endings on the screen as they would be printed. Of the early PC products, we were the only one to do this. As time went on, and especially after the introduction of the Macintosh, users wanted a "true" WYSIWYG, with fonts in the same size and style as they were to be printed, with headers, footers, and footnotes in the right places, and with any figures included in the document accurately displayed on the screen.
Offering a true WYSIWYG word processor for DOS was not an easy thing to do. DOS was designed to work in either a text mode or a graphics mode, but the screen display was much slower in a graphics mode. In a text mode, the computer handled the screen as if it had 25 lines of text with up to 80 characters on a line. This amounted to only 2,000 boxes on the screen (25 times 80) for the program to worry about. In a graphics mode, the computer worked with tiny dots. With the standard 640 times 480 dots for every screen, this amounted to literally hundreds of thousands of dots for the computer to address. Since hundreds of thousands of dots takes much longer than handling two thousand boxes, the screen display in a graphics mode is much slower.
Another problem with a graphics mode was that DOS came with only the barest of essentials for controlling the screen. To control the screen properly to handle the fonts and graphics, we either had to go to the trouble to write our own screen drivers and come up with our own screen fonts, or try to use Windows. Windows was designed to provide these services, but it was still slow and unreliable. Microsoft needed more time and customers needed faster computers before Windows would be an answer to the WYSIWYG question.
After long discussions about whether to produce a full WYSIWYG version with DOS, we decided on a compromise. Normal typing would be done as always in a text mode. A graphics mode would be used only for doing a preview of the printed page and for working with the graphics figures. A print preview was not as good as WYSIWYG editing, but it did give the writer a good view of how the document would look when printed. This compromise saved us a lot of work and was probably the best solution for the customer until faster machines came along. Full WYSIWYG would have to wait until 6.0.
Besides improvements to printing and the addition of graphics, there were a lot of new features planned for 5.0. The major features included automatic referencing (if "see page 14" appeared in a document and then a few pages were added, "see page 14" would automatically change to "see page 17"); support for more than 1500 characters including international characters and diacriticals; intelligent printing (if a document designed for one printer was moved to another printer with a different set of fonts, WP would decide which of the new fonts to use and how best to print the document); master documents (many documents could be combined and printed together as one document); bulleted outlines (lists with bullets in front of each item); and automatic redline and strikeout. There were hundreds of other minor features as well.
We spent a lot of time in meetings going over what had to be in the product and how things should work. These decisions were made by the programmers, who sometimes had very heated discussions about what was needed. At times the three of us on the Board had to assume the role of referee. Some of the issues were very complicated, so by the time the arguments were finished and a decision was made, I usually had a headache.
It was somewhat unusual for a software company to let the programmers decide the future of its products. We were, however, a company founded and owned by programmers, where programmers were treated with an extra measure of respect. The marketing department was used primarily to sell products once they were developed, and only rarely did it get involved early enough to perform the traditional marketing role of identifying a need and defining a product to fill that need. At times this put us in the position of developing solutions before we identified problems, but it was hard to be too critical of the programmers when the company was so successful. To their credit, the programmers tried very hard to listen to our customers and to those of us in the marketing department. The programmers were smart and thoughtful and very good at protecting the best interests of the company. At times, however, they were prone to manipulate some of the data they received to fit what it was they wanted to do.
Although much of the work we did that year was very hard, we did have some fun in 1987. At one of our high level meetings (this was the name we gave to the meetings Alan liked to hold on the ski slopes), we talked about the new laptop computers. As we rode the lifts, we came up with the idea for Executive WordPerfect, which would be a collection of software that included junior-type versions of WordPerfect and MathPlan, along with the calculator, shell, and notebook from Library. The product, which was conceived in February, was shipped in May. Executive actually sold fairly well and gave us something new to talk about at Spring COMDEX in Atlanta.
We also came up with a knock out punch for WordStar that year. Since we had Micropro on the ropes, we thought we should put them away for good. We commissioned two editors from one of the best trade publications to write a book entitled A WordStar User's Survivor's Guide to WordPerfect. The two writers, who had used WordStar for many years and liked it, wrote the book under the pen names W. P. Forever and W. S. Farewell. Not only was the book funny and useful, but we now had two more press people that knew our product inside and out.
We built an advertising campaign aimed at diehard WordStar users, offering the book to them for a dollar along with a discount coupon for WordPerfect. We actually gave most of the copies away for free, because the cost to run a credit card charge through our fulfillment company was more than the dollar we were charging. The only reason we asked for a little money was to scare off those people who did not really need the book. The campaign was a great success, unless of course, you happened to work at Micropro.
By now it had been almost a year since I first tried to write our company's purpose, goals, and objectives. At first I had a lot of trouble trying to figure out the difference between goals and objectives. The dictionary I looked in was not a lot of help, using each word in the definition of the other. After agonizing over the definitions for a few months, I finally came to the conclusion that a goal was a desired result you could measure, while an objective was something desired that was easier to visualize or describe than to measure. By this definition, reaching $100 million in sales or scoring a touchdown were goals, because they involv ed reaching a particular number or crossing a certain line. Offering the best product or being honest in every business dealing became objectives, because there was no number or line involved. Likewise, losing ten pounds would be a goal, but looking good in clothes would be an objective.
I was not sure that it mattered whether or not I understood these words correctly, but the exercise of trying to understand them helped. I had never liked the idea of assigning a number to every goal or objective. Being "fair and honest" did not have to be translated into "less than 15 complaints to the Better Business Bureau." "Fair compensation" did not have to be defined as "an employee turnover rate of 4% per year or less." "Working hard" did not have to mean reaching a sales goal. I liked using images and what images suggested, rather than precise measurements. If we all had an image or an idea in our minds of what it was we hoped to accomplish, I felt we would have a greater chance of success than if we lived and died by our numbers.
Sometime in the spring Alan went to a one day seminar and came back with our mission statement. I was more than a little frustrated with him. I had spent most of a year getting ready to suggest our purpose and objectives, and he had written them down in an afternoon. The definition of our principles seemed so important to me and so central to the proper management of the company that I wanted the definitions to be perfect.
Alan's mission statement gave me the incentive to finally get moving. I took my work, incorporated many of Alan's ideas, talked to Bruce, and by May published the first version of our purpose and objectives. Our purpose was to develop, market, and support the finest software in the world. Our objectives were to maintain the highest standards of honesty, quality, and service; to maintain a management structure which was efficient and rewarding based on team proficiency; to maintain a qualify of life which would encourage employees to stay with the company, keeping turnover to a minimum; and to avoid indebtedness or any public stock offering.
I also put together a long list of policies and procedures, which with our purpose and objectives I called our principles, but I did not pass them around very much. We were still a lawless community of superstars who were not ready for any serious rules.
In June we went to PC/Expo knowing Microsoft was going to show Word 4.0 for the first time. We spent a lot of time in their booth examining the new product very carefully, looking for weaknesses and comparing its features to what we had coming in 5.0. They had improved the speed of the program and added a faster moving cursor. They had included a WordPerfect-like macro feature with a macro language and a redline/strikeout mode as well as some file management features and screen options so users could make their Word screens look a lot more like our WordPerfect screen.
After hearing so many rumors about Microsoft's new product for a number of months, we were all relieved to finally see it. André summed it up best when he said, "Word is still Word." Though they had some interesting new features, we were in good shape. Their interface was still confusing and their product was still from the old run-off/repagination school of thought.
The best way to understand how our products were different back then was to compare Word and WordPerfect to a magic trick. In our case, we understood that word processing was an illusion. The computer was never designed for writing, and it was only by magic that a computer screen could look like the printed page. In their case, Microsoft seemed to use brute force to get the job done. Their columns were inflexible, and they had to use a repagination before printing. Anything fancy had to be done with a style. It was as if they were not trying to hide the tricks to their magic. If we had both been doing the old magic trick of sawing a lady in half, we would have known enough not to hurt the woman when creating the illusion, but Microsoft would have cut her in half and sewed her back up again. Perhaps I overstate our advantages, but not by much. The only time I saw much good magic come out of Microsoft was with Excel (a GUI spreadsheet). I hoped for our sake that Excel was merely an accident.
Not everyone in the press shared our view of Word 4.0, because it received some good reviews. By the end of the summer we were tired of reading stories about Word's terrific laser printing. We were tired of their promotions, their advertisements, and their slams against WordPerfect. We were tired of hearing they would soon be number one.
On Labor Day a few of us in the marketing department were working even though it was a holiday. About six of us went to a long lunch and talked about how fed up we were with listening to Microsoft's claims of a better product. We were in an especially bad mood, because we had recently learned that 5.0 would not be ready for COMDEX in the fall. The release date had been pushed back to February or March, and we were not looking forward to letting Microsoft beat us up for another six months. The longer we talked, the angrier we got, until finally we decided we were going to fight back. The six of us at the table put our hands together, much like a basketball team at the start of the big game, and said aloud that we were making a formal declaration of war against Microsoft. These were desperate times, and we were going to take desperate measures. Instead of maintaining our normal nice and polite manner, we were going to dispute their false claims, point out their weaknesses, and, if necessary, stretch a few rules.
Stretching the rules did not mean doing anything illegal. It meant we would systematically leak information about 5.0 to the press to build interest in the product. Although pre-announcing the product was likely to hurt our sales a little, it was also likely to convince potential buyers of Microsoft Word to postpone their purchases until our product was released. This was a little out of character for us, but this was war.
The first step in our battle campaign was to hold a marketing boot camp. For the first time we had formal training for all salespeople who worked at the home office. The training stressed teamwork and planning, both of which were necessary if we were to win against Microsoft. For the next few months, each group was to come up with a plan and get approval for it. Each plan was to include a list of objectives and the responsibilities for each person in the group. Everyone was to know what it was they were supposed to do, and the plans were to be shared so that everyone would know what everyone else was doing. I stressed that "responsibility was given without autonomy." In other words, everyone was expected to communicate what it was they were doing with others, so others could offer suggestions and improvements. We were not to be a group of individuals, each working privately and independently. We were a team and we were united against our opponent.
Our second step was to pre-announce WordPerfect 5.0. Instead of waiting until our product was ready to ship or in a beta test, we starting leaking information right away. All of us who were traveling started showing the product "confidentially" to dealers and large accounts. André showed the product to standing room only crowds at Softeach, a series of dealer seminars put on by Softsel. I started a series of messages which I wrote and posted each night on CompuServe. I called the messages "Bedtime Stories," because I posted one message each night just before I went to sleep. For the next two months, I provided a detailed description of 5.0 one page at a time. At first I revealed very little, giving only the history of WordPerfect Corporation and some of our objectives for the new product, and then, much like a striptease, I revealed the 5.0 secrets one at a time. As I introduced each new feature, I explained in detail how it would work. I also asked suggestions and used the forum to conduct last minute market research. The stories were monitored by the press and widely republished. They did a great job of creating interest in 5.0.
Next, we brought in our 35 reps from around the country for the week of October 26 and showed them 5.0. Even though we were still adding features to the product, we sent them home with demonstration copies of the software. Starting November 2, all of them would be able to show their customers and dealers what we were showing at COMDEX. That same week we visited many of the trade publications and let them have preview copies of the software, so they could write about 5.0 in their COMDEX issues. Although the early announcement put a lot more pressure on our programmers to get the product ready and gave Microsoft more time to copy what we were doing, we felt we had no choice.
Anyone who thinks we might have been overreacting to Microsoft's threat to our future sales has likely never run a business. Our free enterprise system is not designed for the squeamish. It is based on competition and on the survival of the fittest. While on the surface capitalism is admirable because it offers everyone an equal opportunity to prosper, there is an ugly side to it. When many businesses go after the same customer, and especially when some of those businesses are desperate to make a sale, there is a fierceness to the competition which is as frightening and violent as any fight to the death in the jungle. Laws which are supposed to prevent the strong from preying on the weak do not help much. Although competitors are almost always civil and courteous to one another, at the same time they are at each other's throats.
At COMDEX we held a press conference to formally announce 5.0 and we showed the unfinished product in our booth. André and his staff had put together another amazing show. With the help of our actor friends from Chicago, they created a classroom for word processing pilots called Top One. Those attending listened to music from the movie Top Gun as they took their seats, put on their Top One hats, and became student pilots. During the show they earned their Top One wings by watching our actor-teachers demonstrate all the new and amazing features in 5.0. In an "officers' club" we gave special attention to our dealers and large customers. André's work was so good that year that it earned him an award from Sizzle Magazine for the best trade show presentation of 1987.
Our Mac group was showing its product as well. The programmers were slowly getting rid of their bugs, but their release date was still up in the air. They wanted to be out before the end of the year, but would not make it. At the Christmas party, the Mac developers came with grocery bags over their heads, singing "Waiting for the Macintosh to Ship" to the tune of "Walking in a Winter Wonderland" to hide their embarrassment.
While our 5.0 news was big, the biggest news at COMDEX in 1987 was OS/2 and Presentations Manager. IBM and Microsoft had previously announced their intention to work together on what was to be the next major PC operating system. Named OS/2, the operating system included the Presentation Manager, which, like Windows, was a platform for GUI applications. For reasons known only to IBM and Microsoft, Presentation Manager and Windows were different, so applications companies like us had to write two versions of their products in order to support both platforms.
I participated on an OS/2 panel at the show and was stunned by the audience and its feelings about OS/2. Except for a very small minority in the audience, everyone there was interested in buying and using OS/2 with the Presentation Manager. There was virtually no interest in any applications written for any other operating system, including those written for Windows. No one seemed to have a clear idea as to why they wanted OS/2 PM, but they all seemed to believe there would be many new and wonderful applications and many new computing improvements resulting from the new operating system. Customers were expecting the unimaginable.
I felt like an outsider. I had no clue as to what the customers were hoping for, and we were too busy worrying about winning the DOS market to get too excited about OS/2. It disturbed me to see people so excited about an operating system which was unavailable. Nevertheless, the race was on for software developers to come up with the next 1-2-3 or the next application which would define the new platform and compel customers to buy OS/2 PM.
As I spoke on the panel, one member of the audience yelled out, "Don't waste your time on the Amiga," which resulted in many people clapping in agreement. In retrospect, the hecklers were probably right. We should not have wasted our time on the Amiga, but we were not ready to put 50% of our programmers on an OS/2 PM project either. DOS was going to be an important operating system for a long time, and many of our other WordPerfect versions were bringing in significant revenues. We would start work on OS/2, but we would not drop everything else to do it.
As we neared the end of the year, it looked like sales would end up at $97 or $98 million. This was the worst possible outcome. We were too close to the $100 million goal not to make it, so Bruce and Alan asked the marketing department to come up with some sales. We called a few of our distributors and asked them for help, and as it turned out, they were expecting our calls. They were hoping we would be anxious enough to make the goal to offer them an extra discount or better terms. We did in fact offer them special terms, and as part of the deal asked them to place their orders by telegram.
The telegrams arrived the day of our company party. As was the custom, Alan, Bruce, and I each spoke for a few minutes. Alan started his speech by announcing our current sales. As he talked, we delivered the telegrams to him one by one. With each telegram he gave the new sales total. When the last telegram was delivered and read, Alan dramatically announced that we had reached our goal. We ended the year with sales of $100,350,000, just enough to send everyone on vacation in the middle of our war with Microsoft.
1987 ended with everyone in the company on their feet cheering, clapping, and screaming, and that is how it should have been. The year was full of firsts. On Softsel's October 26th hot list we had made it to the top for all personal computer software products, ahead of 1-2-3 and dBase III. Though Microsoft Word was fourth, our WordPerfect network station pack was listed eleventh. Amazingly, Library was listed as number 12--the first time we had more than one product on Softsel's hot list. We also finished the first of many new buildings in the research park. Our pre-tax profit margin was an enviable 33%, and we were making serious enough money to pay for the buildings as we built them. It also looked as if we were winning our war with Microsoft.