Editor's note: In November 2025, I published an episode of Chronicles Revisited Podcast on the history of Software Publishing Corporation. Below is a lightly edited version of my recording script for that episode.
Introduction
Welcome to the Chronicles Revisited Podcast. I’m S.M. Oliva. I write the blog “Computer Chronicles Revisited,” which reviews the people, products, and companies featured on the PBS series that aired between 1983 and 2002. In this podcast, I go in-depth on stories previously featured on the blog.
For this episode, I’m looking at Software Publishing Corporation. Formed in 1980, SPC was an early leader in the market for personal computer—or as it was then more commonly called, micro-computer—software. Despite the name, Software Publishing was first and foremost a developer. Its initial "PFS" product line offered an inexpensive, modular, and easy-to-use approach to productivity software for a number of early microcomputer platforms such as the Apple II and the IBM Personal Computer.
But as the microcomputer market started to consolidate behind the IBM PC standard, Software Publishing shifted focus, first to more complex application software and then to a professional-grade business presentation program, Harvard Graphics, which came to define the latter half of the company's existence. After hitting its peak in terms of sales and profits in the early 1990s, by the middle of the decade Software Publishing found itself struggling to complete in an ever-more complicated world of application software dominated by Microsoft. This led to Software Publishing's acquisition by an upstart "new media" company that itself disappeared into the ether of the dot-com bust of the early 2000s.
John Page's Transition from UK Hardware to US Software
While Software Publishing Corporation developed software for personal computers, its three co-founders were veterans of Hewlett-Packard’s mini-computer division. Let's start with the person responsible for SPC's earliest software products, John Page.
Born in 1944 in the West Kensington area of London, John Page grew up in the post-war austerity of Great Britain. Page's father owned a small shoe repair business while his mother worked as a seamstress and housecleaner. Page himself attended a boarding school due to a policy implemented by the post-war Labour government requiring private schools to accept a number of students for free each year based on exam results.
Page, however, felt isolated from the other kids at school due to his working class background. In his own words, he was a "reclusive nerd." This led him to develop an interest in electronics, specifically amateur radio, and during his teenage years he built his own receivers using money he earned delivering shoes for his father.
Sadly, by 1961, his father's health had started to decline and the now-17-year-old Page needed a job to help his family. Through a friend of his father, Page secured an apprenticeship with LEO, the computer division of British conglomerate J. Lyons & Company.
LEO, which stood for Lyons Electric Office, manufactured a series of business mainframes modeled on the EDSAC, one of the earliest digital computers. When Page arrived, the LEO mainframes had just completed the transition from vacuum tubes to transistors. And one of new apprentice Page's jobs was counting those transistors as they arrived in boxes.
After moving around the different departments of LEO performing similar low-level tasks over the next couple of years, one day Page went to his floor manager and asked if he could work on the actual computers. According to Page, the manager replied, "I don't see why not," and immediately took him to see one of the senior engineers, who took Page under his wing as an assistant.
Page subsequently learned how to build and maintain LEO III mainframes for the company's clients. One of the last such machines that Page worked on was in 1968 for the government of Czechoslovakia. While debugging the computer on-site in Prague, Page's boss called and ordered him to come home right away. Page ignored the order as he wanted to finish the job. Eventually, Page's boss showed up in-person and said, "Come on, we're leaving." On the flight back to London, Page's boss explained the reason for their hasty exit: "The Russians are going to invade Prague tomorrow." Tomorrow was August 20, 1968. That day, the Soviet Union and three of its its Warsaw Pact satellite states invaded Czechoslovakia in a successful effort to silence anti-government protesters and strengthen the country's Communist government.
Meanwhile, back in the United Kingdom, Prime Minister Harold Wilson's Labour government passed the Industrial Expansion Act, which led to the creation of a new state-backed computer monopoly, International Computers Limited or ICL. Most of the UK's existing computer companies, including LEO, were merged into ICL. This consolidation led to the closure of the LEO factory where John Page worked and he was laid off.
After a brief stint working for a company building factory automation systems, in 1970 Page answered a newspaper ad seeking "people with computer knowledge" for a job at Hewlett-Packard's UK office in the Berkshire town of Slough. The job turned out to be in sales, not engineering, but Page still jumped at the opportunity. The Slough office needed someone who knew enough about computers to help the salespeople meet their quotas. When Page described his own background working on the LEO III, he was hired on the spot.
Page initially worked in sales support. At the time, HP's main computer line was its 2100 series of 16-bit mini-computers. To give you some idea of what these computers were like, the initial model, the 2116A, debuted in 1966, came with 8 kilobytes of RAM, and cost $22,000.
In 1972, HP started rolling out a higher-end mini-computer line, the HP 3000. The original 3000 was a market failure that the company quickly withdrew from the market. But in 1974, a revised model, the 3000 CX, proved an instant success. Although, success is relative when we're talking mini-computers. HP received a total of 20 orders for the CX during its first three months on the market. But at roughly $100,000 per machine, that came to $2 million in revenue.
To sell the revised 3000 line, HP needed someone from the UK office to go to Geneva, Switzerland, for a year to run a showroom. Page volunteered. This first required him to travel to HP's headquarters in Cupertino, California, where he'd spend the next four months learning about the 3000 from its development team.
After completing his trek from London to California to Geneva and then back to London, Page got a call from HP HQ asking if he'd be interested in coming back to Cupertino to run the worldwide support organization for the 3000. He said yes and returned to the United States in 1975. Of course, the job was still sales and customer support, which weren't Page's real interests. But he managed to work his way over to the software side of the business, eventually overseeing the team that developed "Image," the database management system bundled with the 3000 lijne.
Fred Gibbons Has a Micro Epiphany
This brings us to late 1979. One day at the HP office, Page ran into a colleague, Fred Gibbons, one of the hardware marketing managers for the 3000. Gibbons asked Page if he'd heard about VisiCalc, the spreadsheet application recently released for the Apple II microcomputer. Page said he'd seen VisiCalc but had a hard time believing that such a small computer could do actual business work.
Gibbons wasn't simply making chit-chat. He'd been working on a business plan to create his own software publishing company—or Software Publishing Corporation, if you will—that would follow VisiCalc's example in selling commercial software for microcomputers.
Fred Gibbons was born in Massachusetts in 1949. Raised in Cape Cod, where his father worked as a commercial sea captain, Gibbons was a natural salesman. As a 17-year-old high school student in the 1960s he started his first business designing and selling canvas bags for skis. Two years later, he acquired a franchise to sell Hobie Cat catamarans on Cape Cod.
Gibbons used his earnings from these early ventures to fund his education at the University of Michigan, where he earned bachelor's and master's degrees in computer science. While Gibbons didn't grow up an electronics enthusiast like John Page—Fred was more the outdoors type—he found computer science attractive because, as he put it, "it's an ordered world, very structured. Programming is easy, it's rote, it's either right or wrong."
After finishing his master's degree, Gibbons returned to Cape Cod and spent a year working for Stewart Systems, a small local computer manufacturer. Then he decided to bolster his management credentials by getting an MBA at Harvard University in Cambridge. After graduating in 1975, Gibbons joined Hewlett-Packard and quickly rose through the ranks, reportedly becoming the company's youngest-ever marketing manager.
One day in 1979, Gibbons received a recruiting call from Steve Jobs, the co-founder and then vice president of marketing at Apple Computer. The success of VisiCalc had proven a significant boon to Apple. While the Apple II, designed by Jobs' co-founder (and former HP engineer) Steve Wozniak, had been on the market since 1977, it wasn't until VisiCalc that business customers started taking the machine seriously. Apple planned to capitalize on this by releasing a new business-focused microcomputer, the Apple III. And Jobs wanted Gibbons to lead the charge as the Apple III's marketing manager.
Gibbons declined the offer. He later admitted, "Steve is a tough guy to turn down," but at the time he wasn't looking to leave Hewlett-Packard. Of course, in hindsight that was a smart move as the Apple III, released at the end of 1980, proved an even bigger bust than the original HP 3000. At the same time, however, Jobs' invitation got Gibbons thinking about the state of the microcomputer market.
So Gibbons decided to do what he knew best: market research. He visited some computer stores near the HP offices and was struck not so much by the computers but by the standalone software programs for sale. This was a foreign concept to someone who sold minicomputers for a living. For example, a typical HP 3000 configuration sold for about $70,000 in 1979. But that included all of the basic application software, including the Image database developed by John Page's group.
This wasn't the type of software you could just walk into a store and buy off the shelf. Nor was it something you could easily install, learn, and use on your own. Indeed, HP offered Image training at $500 per person, or $4,500 for a five-day on-site course.
In contrast, VisiCalc sold for about $100 at launch. Of course, you needed an Apple II, which would set you back around $2,300 for a complete system, but that was still a bargain compared to a minicomputer. More to the point, the Apple II was easily expandable with off-the-shelf hardware and software that didn't require a multi-thousand-dollar training course.
Based on his market research, Gibbons started drawing up a business plan to get into what he saw as a still largely untapped market for microcomputer application software. He was particularly interested in developing a database, as there wasn't one yet on the market for micros. (Ashton-Tate's dBase, which I covered in Episode 9 of this podcast, wasn't out yet.) This was what prompted Gibbons to approach John Page.
Page was intrigued by the prospect of writing a database for microcomputers. He'd been increasingly frustrated by how the database systems he developed for HP's minicomputers was tailored to programmers and managers rather than end users. He saw the microcomputer as a chance to create a database for ordinary people.
So Page agreed to work on the idea part-time during his off hours. Gibbons provided an Apple II for Page to try out. Page was nervous about the technical challenge. As he saw it, Gibbons had effectively asked him to cram a quart of programming into a pint-sized container. But Page reasoned that the VisiCalc guys managed to pull it off, so he could as well.
A few days into his work, Page decided to make his own pilgrimage to a local computer store for inspiration. Initially, he wasn't encouraged by what he saw. Most of the computers didn't even come with a disk drive, instead relying on cassette tapes to store programs. As far as he was concerned, these were toys, not serious computers.
Still, as he'd later recall, Page saw the micro-computer as the "beginnings of something big," so he continued to develop his database program. He decided that the only micro-computer platform worth targeting was the Apple II. There were two key reasons. First, the Apple II had a disk drive, one developed by Steve Wozniak himself. Second, while a base Apple II model came with just 16 kilobytes of RAM, you could also purchase configurations with 32 or 48 kilobytes. Page knew that 16 or 32 kilobytes were non-starters, even for a simple database program, but he could make 48 kilobytes work.
Page's next technical decision was how to program the database. By default the Apple II only had two choices of programming languages: Applesoft BASIC and 6502 assembly language. The former wouldn't work, as BASIC was not suited for developing advanced applications. And Page felt it would take too long to write out his database program in assembly. He needed a third option.
Fortunately, in August 1979 Apple released Apple Pascal, an implementation of the Pascal programming language originally developed at the University of California, San Diego. Pascal was still an interpreted language like BASIC, which was slower and less ideal for Page's needs than a compiled language like C. But C wasn't an option on a 8-bit microcomputer. So Page went with Pascal, although he would end up re-coding several portions his draft program in assembly to improve performance.
Janelle Bedke Brings It All Together
With John Page working on the product, Fred Gibbons now had to figure out how to market it. He initially approached his bosses at Hewlett-Packard, but they weren't ready to get into the microcomputer business in 1979. (If you want to know more about when HP did finally make the move into micros, listen to episode 14 of this podcast.) Gibbons also considered working through an existing publisher, but there weren't that many around at the time. So the next step was naturally to start his own publishing company.
Now, the name Software Publishing Corporation would end up being something of a misnomer. Originally, Gibbons thought he would simply pay John Page a royalty to publish his database program once it was completed. Gibbons imagined SPC as the equivalent of a music publisher selling records for a record player. But once SPC was actually up and running, the royalty plan fell by the wayside and Page became a co-founder and later vice president in charge of an in-house software development department.
Page was not the only co-founder that Gibbons recruited. To complete what ended up being a triumvirate, Gibbons approached yet another colleague in HP's minicomputer division, Janelle Bedke, to join him and Page.
Janelle Bedke was born in Sacramento, California, in 1948. She was the third of four daughters of Colonel Hazen Bedke, a longtime meteorologist with the United States Air Force. Due to their father's career, Janelle and her sisters would end up moving between California, Washington, DC, Germany, and Hawaii, before finally settling in Salt Lake City, Utah, where Colonel Bedke headed the western regional office of the National Weather Service after he retired from the military.
The colonel strongly encouraged all of his daughters to go to college and pursue professional careers. Janelle herself demonstrated an early aptitude for mathematics. Despite living the life of an "Air Force brat" and changing schools constantly, Bedke graduated from high school at the age of 16 and enrolled at her father's alma mater, the University of Utah, in 1965.
By her sophomore year, Bedke realized she had little interest in pursuing a career either teaching mathematics or conducting academic research. She was looking for something more practical. So she decided to pursue computer science as a specialty. Now, in the late 1960s, Utah didn't offer an undergraduate major in computer science. Bedke simply took whatever computer science and engineering classes she could.
And it paid off. After graduating magna cum laude from Utah in 1969, the now-20-year-old Bedke interviewed with a number of companies before deciding to join Hewlett-Packard. Bedke would later recall that when she visited HP's offices in Palo Alto, California, it was the first time she was able to touch an actual mini-computer, as Utah only had larger mainframes accessible by dumb terminals.
During her first five years at HP, Bedke worked in software development for the company's minicomputer division. In 1974, she was promoted to an R&D project manager, were she ran a lab developing a line of microprocessor-based terminals. These terminals were essentially an intermediary step between dumb terminals and fully-fledged microcomputers. Bedke later recalled this was an especially exciting development as it meant computers would become "more and more usable by more people."
But by 1979, working in middle management started to chafe at Bedke. In an interview for a 1987 book on women in technology, Bedke said, "HP got to be a large company and it was difficult to see any direct results. Things bubble up into the system, and I was never sure I had made a difference. It was frustrating."
These frustrations made Bedke receptive to Fred Gibbons' offer to join his proposed startup as general manager and head of operations. From his perspective, Gibbons said Bedke was "an outstanding doer," that is, someone who had complete command of the technical and execution details of getting a product to market.
With the three co-founders assembled, Software Publishing Corporation was formally incorporated on April 18, 1980. That didn't mean that SPC had a product ready to go. It would take John Page another six months to finish his database. In the meantime, Janelle Bedke got to work. She was the first of the trio to quit her job at Hewlett-Packard. She'd spend the next several months laying the groundwork for SPC's public launch from her home in Atherton, California. Fred Gibbons put up his personal savings of $50,000 as SPC's startup capital. And both he and Page kept working at HP, giving a portion of their paychecks each week to Bedke to keep her going.
Finally, in September 1980, Software Publishing Corporation released John Page's Apple II database program, which was named pfs:FILE. PFS officially stood for "Personal Filing System," but in practice it was simply a prefix that SPC would use to describe its initial series of application programs.
To publicly mark the release of PFS FILE, SPC took out the back-page ad in Softalk, itself a recently launched magazine focused on Apple II users. The ad copy proclaimed, "Using PFS at work you can make better decisions by creating files on inventory, customers, or orders and accessing them in seconds." And running on a 48K Apple II, the $95 pfs:File could store up to 1,000 records on a single five-and-a-quarter-inch floppy disk.
Not long after Softalk hit the newsstands and Janelle Bedke sent a letter to hundreds of authorized Apple II dealers announcing the launch, she received her first telephone order from a computer store in San Diego. SPC still existed entirely in Bedke's house at this point, so for the first few months she was filling orders from her dining room and garage. At one point, she was even called down to the local police station in Atherton after her neighbors complained about the constant presence of delivery trucks in front of her house.
Of course, that initial burst of activity was a good sign for SPC. pfs:File proved an instant commercial and critical success. Softalk's own review proclaimed John Page's work "a quantum leap forward in data base programs, analogous in achievement to VisiCalc." And units sales, while measured in three figures during these early months, were good enough to prompt Fred Gibbons and John Page to leave Hewlett-Packard for good in early 1981 and assume full-time roles at SPC.
Expanding the Lineup
In May 1981, SPC finally moved out of Janelle Bedke's house and into proper offices in Mountain View, California. A few weeks later, on June 10, 1981, the company launched its second product, pfs:Report. This was essentially an add-on module for pfs:File. Report, as the name implied, simply added the ability to create reports from information in a pfs:File database. But due to the Apple II's 48 KB memory restriction, John Page decided he couldn't fit reporting features into pfs:File, hence the separate program.
This would end up playing into SPC's marketing approach as the PFS series grew. There would ultimately be seven titles released in the initial series, each a standalone module retailing for between $100 and $150. All of the programs would share a common user interface and design language that emphasized simplicity and ease of use. SPC's advertising emphasized this with the slogan: "The Power of Simplicity."
The launch of pfs:Report helped give SPC its first $100,000 sales month in June 1981. John Page was able to hire additional programmers to assist him with remaining five PFS programs, while Janelle Bedke brought in a full-time writer to take over creating the product manuals, something that Bedke had been doing herself. The company also reached a deal to sell 30,000 units of pfs:File directly to Apple for use in a 1981 holiday promotion.
By the end of SPC's first full fiscal year in September 1981, the company reported $96,000 in profits on $732,000 in revenue. A year later, that figure jumped to just over $4 million in sales and $627,000 in profits. By September 1983, sales reached $10.2 million with $1.2 million in profits.
In less than three full years, SPC grew from a three-person startup to an established company with 79 employees. The company now had four products, adding pfs:Graph and pfs:Write to its complement pfs:File and pfs:Report. The PFS series was also no longer exclusive to the Apple II.
Fred Gibbons had negotiated a deal with IBM to produce and sell proprietary editions of the PFS programs for the IBM Personal Computer. Known as the "Assistant Series," these were modified versions of pfs:File and the other applications that conformed to IBM's interface standards and optimized the PC's peripherals. SPC also negotiated OEM agreements to bundle PFS software with several other computers, including machines from Panasonic, Tandy-Radio Shack, Digital Equipment Corporation, and even Hewlett-Packard, which had finally entered the micro-computer business.
Another initiative Gibbons launched in 1983 was a mail-order catalog division called Power Up!, which initially sold 40 software packages and hardware accessories for IBM PC and Apple II computers direct to consumers. The catalog did not include any of the PFS programs. Instead, Gibbons' wanted to market other software titles that were unlikely to see wide distribution through traditional computer stores.
As SPC entered 1984, Fred Gibbons and his co-founders prepared to take their company public. While the trio had boot-strapped the launch of PFS FILE back in 1981, that initial product's quick success prompted Gibbons to seek two rounds of venture capital funding. And on November 15, 1984, SPC made its debut as a public company with an initial offering of 1.6 million shares. About one-fifth of those shares were the venture capitalists cashing out, but SPC's haul still came to about $10 million.
Nineteen eighty-four also marked the completion of the original series of seven PFS products. The final programs introduced that year were pfs:Proof, a spell checker; pfs:Access, a communications package to access on-line database services; and pfs:Plan, a spreadsheet. Coincidentally, Paul Schindler reviewed all three of these programs for Computer Chronicles during the show's second and third seasons.
The continued critical and commercial success of the PFS series led to another record year for Software Publishing. For the fiscal year ending September 30, 1985, the company's first as a public company, SPC posted $37.2 million in sales and just over $5.8 million in profit.
Nineteen eighty-five also marked the start of a transitional period for Software Publishing. There were a couple of different factors at play. First, there was an industry-wide slowdown in new hardware sales. This mostly affected lower-end and home computers as opposed to business machines like the IBM PC and its multitude of clones. But it effectively marked the end of many other computer lines that struggled to remain relevant in the wake of the emerging PC standard.
SPC's approach to product development during its first four years had been built around Fred Gibbons' belief that software was no different than music running on a music player. Just as a record label supported every format, such as vinyl, cassette tape, and compact disc, so too did SPC support just about every personal computer platform on the market.
Eventually, this led to what John Page later called a "crisis," as the company started to miss ship dates as the number of stock-keeping units continued to multiply. Page confronted Gibbons and implored him to cut back on the number of supported platforms. Unfortunately, it was at that moment Steve Jobs appeared at SPC's Mountain View offices and demanded Gibbons support his newest platform: the Apple Macintosh.
The Macintosh was Apple's third attempt to build a business computer following the unqualified failures of the Apple III and the Apple Lisa. Jobs desperately needed rock-solid third-party support to make his Mac viable. To that end, he convinced Gibbons, along with Microsoft's Bill Gates and Lotus Development Corporation's Mitchell Kapor, to participate in a bizarre launch event themed around the 1970s game show "The Dating Game":
Yielding to Jobs' pressure, Gibbons convinced Page to release pfs:File on the Macintosh. This proved to be a failure. Page said the problem was that Gibbons assumed this would just be like every other port they had done. But the Macintosh was a completely different architecture from the IBM PC and the 8-bit machines that had been SPC's bread and butter. SPC's products simply weren't built with the Macintosh's tight hardware-software integration and graphical user interface in mind.
Meanwhile, even as Gibbons was trying to help his old friend Steve Jobs by supporting the Macintosh, Apple was cutting into sales of SPC's flagship PFS series on the Apple II with a competing package, AppleWorks. Launched in 1984, AppleWorks provided an integrated word processor, spreadsheet, and database for around $250.
Even without competition from Apple, as more companies looked to copy SPC's model by offering lower-price standalone software applications, it was getting harder for Gibbons to deliver the constant growth demanded by stock market speculators. This was underscored by Software Publishing's dismal 1986 results. For the fiscal year ending September 30, SPC reported a 36 percent decline in sales and an 88 percent decline in profits from 1985.
Fred Gibbons' response was swift. In July 1986 SPC launched two new product lines. The first was pfs:First Choice, a $149 integrated software package that included a word processor, spreadsheet, file manager, and communications software. The second was the PFS: Professional Series, which were higher-end and more expensive versions of the standalone PFS applications. In a break with the past, these new products were exclusively for IBM PCs and compatible MS-DOS machines. Gibbons conceded in an interview with Lotus magazine's Paul Freiberger that John Page had been right about SPC being stretched thin supporting too may computer platforms: "[O]ur R&D got slowed going sideways instead of forward. We should have said the world is going to go totally PC-compatible."
The Harvard Re-Branding
As part of his PC epiphany, Gibbons also did something he'd resisted up to this point—buy another software company. In August 1985, Software Publishing Corporation acquired Harvard Software, Inc., in a stock swap valued at around $4.3 million. Despite the name, Harvard Software had no affiliation with the famous university. Rather, it was founded in 1983 in the small town of Harvard, Massachusetts, about 25 miles outside of Boston, although it later relocated to the town of Littleton five miles down the road.
Harvard Software's first product was Harvard Project Manager. Released in November 1983, the $395 project planning package originally targeted construction companies. But Harvard Software soon found it was selling to a wide range of professionals, including real estate developers, personnel evaluators, and even political consultants. This led to the release of a deluxe version, Harvard Total Project Manager, which sold for $495.
After just two years in business, Harvard Software had sales of $3.3 million. But the founders quickly realized that with the software market starting to consolidate in the wake of the industry downturn, running a one-product company wasn't a sustainable long-term operation. Shakeel Mozaffar, Harvard's vice president of marketing and managing director, told the Boston Globe, "We reached a critical threshold of growth that would have required a lot more money and talent." From Harvard's perspective, their best option was to sell to a larger company like Software Publishing Corporation.
For his part, Fred Gibbons told the Globe that he'd "frowned on purchasing established software firms" in the past because he feared a "loss of momentum and staff." But flush with over $20 million in cash following SPC's successful IPO, he decided it was worth taking a risk on Harvard Software. Despite Gibbons' public image as a nice guy, he wasted no time firing Harvard's entire 30-person staff and shuttering its Littleton offices once the deal closed. According to tech journalist (and Computer Chronicles contributor) Wendy Woods, only one now-former Harvard employee accepted an offer to relocate to SPC's offices in California.
Gibbons also took this opportunity to divest the Power Up! catalog division he'd started just two years earlier. In September 1985, Software Publishing sold all of Power Up's assets to a newly created company called Channelmark Corporation. Ed Lauing, the SPC marketing executive who'd been running the division, stayed on as Channelmark's new president and CEO. SPC only gained about $315,000 in the deal. Indeed, the catalog had never turned a profit and its emphasis on low-volume, low-cost software no longer fit into Fred Gibbons' reformed vision of SPC as a developer of premium business software for the IBM PC.
That vision centered on using the newly acquired "Harvard" brand to establish a product that would distinguish—and ultimately distance—SPC from its entry-level PFS brand. Just six months after acquiring Harvard Software, in February 1986 SPC announced Harvard Presentation Graphics, a $395 package for the IBM PC designed to produce graphs, text charts, and organizational charts for use on paper, overhead projector transparencies, and 35-millimeter slides.
Now, presentation graphics software was hardly an original concept. Lotus 1-2-3, the PC industry-standard spreadsheet released back in 1983, had the ability to produce simple charts. But it was not until IBM released its enhanced graphics adapter (EGA) in 1984 that the established application software developers started to take the idea of business graphics more seriously. Companies including Ashton-Tate, Microsoft, and Lotus rushed to market with new chart- and slide-making products targeting corporate managers eager to make their business presentations more exciting.
Notably, all of these companies entered the market through acquisitions. Lotus Development Corporation jumped to an early lead with its standalone graphics offering, Freelance Plus, which came via the acquisition of another small Massachusetts software startup, Graphic Communications Incorporated, in June 1986. According to Software Publishing's own internal marketing research, Freelance Plus was the leading business graphics product in 1987 with a market share of 22 percent. Harvard Graphics—SPC dropped "Presentation" from the name in subsequent releases—accounted for 13 percent. That was good enough for second place, but obviously Fred Gibbons wanted his company to be number one.
The way he planned to do that was quite simple: marketing, marketing, and marketing. In fiscal year 1987, Software Publishing increased its marketing budget by 32 percent. The company also transitioned from relying on third-party sales representatives selling through to retailers to having an in-house sales force to market its products directly to business customers. By September 1987, nearly half of SPC's 206 employees worked in sales, marketing, and customer support. A year later, Gibbons further expanded SPC's support operation by acquiring Office Solutions, Inc., a Madison, Wisconsin-based software company that developed a word processor called Office, which had been featured in a March 1987 Computer Chronicles episode.
OfficeWriter was considered a program for "power users," which fit well into the new SPC's vision. But unlike Harvard Software, where Fred Gibbons quickly moved to shutter the company, his plan for Office Solutions was to keep it running in Madison under the name Software Publishing Wisconsin. The Madison office would assume responsibility for continuing development of both pfs:Professional Write and OfficeWriter while also providing the bulk of SPC's support staff. Gibbons told Madison's Capital Times he hoped that proximity to the University of Wisconsin and the lower cost of living in Madison would make Software Publishing Wisconsin more attractive for new software engineers.
Gibbons' Co-Founders Depart SPC
Now, you may have noticed I've been talking a lot about Fred Gibbons and his vision without mentioning his two co-founders. That's because by early 1988, both Janelle Bedke and John Page had left Software Publishing. In the case of Page, I only found a single reference to his departure in a May 1988 InfoWorld article, where writer Rachel Parker reported there had been a "changing of the guard" at the company and while Page remained "focused on dreaming up new products, the company was headed more towards a marketing focus."
As for Janelle Bedke, her departure was also somewhat sudden and not well explained. But before we get to that, let's go back a bit to Christmas Eve 1985. As I mentioned earlier, Fred Gibbons is an avid outdoors enthusiast, especially skiing. Not surprisingly, he spent his Christmas vacation that year on the slopes of the Olympic Valley, California, ski resort currently known as Palisades Tahoe. While on the slopes, he suffered what initial news reports described as a sudden loss of motor control and paralysis on his left side. Subsequent reports offered conflicting information as to exactly what happened, but suffice to say Gibbons was out of commission for a few months.
During this period, Janelle Bedke stepped up from her role as vice president and general manager and temporarily assumed Gibbons' CEO responsibilities. Gibbons returned a few months into 1986, but then in August 1987, he announced he was relinquishing the president and CEO positions to Bedke on a permanent basis. Gibbons would remain as SPC's chairman and chief strategist.
Yet four months later, just before Christmas 1987, Bedke abruptly resigned and left SPC for good. Press statements at the time only said she left for unspecified "personal reasons." And from my research, I couldn't find a single statement from Bedke in the 38 years since her resignation offering any further explanation. As far as I can tell, she simply retired.
The only hint of what happened comes from an interview that Bedke gave to Computerworld just a week before her unexpected resignation. In outlining her agenda for Software Publishing, Bedke said one of her top priorities was returning to the Macintosh, a platform the company had abandoned after its ports of PFS for the original Mac failed in the market. Bedke told Computerworld, "Some of our products are just begging for the Mac interface," and suggested SPC might acquire an existing Mac software line before the end of 1988.
After retaking his CEO's chair, however, Gibbons put a damper on any Mac talk stating, "We will consider the Mac, but only if we have extra resources—right now we have no plans for the Mac."
In fact, with the long-awaited release of IBM's OS/2 operating system at the end of 1987, Gibbons decided that SPC's long-term future was in focusing on OS/2 and its graphical user environment, Presentation Manager, rather than looking to the Mac or even Microsoft's nascent Windows.
This renewed emphasis on IBM and high-end corporate customers was reflected in Gibbons' reshuffling of SPC's organization following the departure of his two co-founders. While Gibbons retook the president and chief executive positions, he gave up the title of chairman. Jack Melchor, one of Software Publishing's initial venture capital backers who still owned a significant stake in the company, became the new chairman. Melchor was, not surprisingly, a former Hewlett-Packard executive, having once run the Palo Alto division that developed the 2116A mini-computer back in the 1960s.
After retiring from HP in 1968, Melchor established himself as one of Silicon Valley's premiere venture capitalists, backing not just Software Publishing but also The Learning Company and Osborne Computer Corporation, serving briefly as chairman of the latter. Prior to taking on a formal role at SPC, Melchor also dabbled in politics, serving as an advisor to both United States President Ronald Reagan and United Kingdom Prime Minister Margaret Thatcher as they sought to make the world safe for Big Business once more.
Back in Mountain View, Fred Gibbons bolstered his own commitment to the IBM way by elevating a former Big Blue executive, Wes Richards, to the role of vice president of sales and marketing, effectively assuming the number-two slot in the organization previously filled by Janelle Bedke. Rounding out the new management trio was Ken Whitaker, a former executive with the mini-computer company Data General, who took over John Page's position as head of research and development.
From Dizzying Highs to Terrifying Lows
By mid-1988, SPC now had four main product lines: pfs:Professional Series, pfs:First Choice, OfficeWriter, and Harvard Graphics. And there was no question that Harvard Graphics was the most important. In April 1988, SPC's group product manager for the Harvard brand, Tess Reynolds, appeared on Computer Chronicles to demonstrate version 2.1 of Harvard Graphics in a segment that also featured its main rival, Lotus Development's Freelance Plus.
Reynolds had put together Software Publishing's aggressive plan to make Harvard Graphics the top product in its category. And it worked. From SPC's downturn in the fiscal year 1986, when it reported $23.6 million in sales, by 1990 that figure had increased nearly seven-fold to $140.6 million. Profits experienced similar growth, from just $700,000 in 1986 to $19.8 million in 1990. SPC's headcount had also more than tripled from 206 employees in 1987 to a peak of 725 by 1991.
But—and you knew there was a but—Software Publishing was now in a position where its success depended almost entirely on a single $500 product. By September 1990, Harvard Graphics (and Harvard Graphics accessories) represented 70 percent of all Software Publishing sales. Its entry-level PFS series, now called pfs:First Choice, only accounted for 14 percent, with the pfs:Professional Series and OfficeWriter making up the remainder.
In January 1991, Fred Gibbons made the decision to get out of the entry-level market altogether. That month, Software Publishing reached an agreement to divest its entire pfs:First Choice line to Cambridge, Massachusetts, based Spinnaker Software Corporation. Now, the history of Spinnaker is its own bespoke tale, one which my friends Alex Smith and Jeff Daum ably covered in episode 162 of their podcast, They Create Worlds. But in brief, Spinnaker was founded in 1982 as a publisher of educational software. By the early 1990s, however, that market had started to get crowded, and Spinnaker co-founder and CEO David Seuss, under pressure from his board, decided to pivot and turn the company into a seller of low-cost software for home and small business users.
The actual deal to transfer the PFS series to Spinnaker proved more complicated than you might think. Rather than Software Publishing simply selling the product line for cash, SPC agreed to accept payment in Spinnaker stock. Specifically, Spinnaker agreed to transfer 1.86 million of its shares—about 16 percent of the company—to Software Publishing. Additionally, Spinnaker would pay SPC ongoing royalties on any future sales of PFS products.
This was all well and good, except that for some reason, Spinnaker never actually delivered the promised shares. This led Software Publishing to file a breach of contract lawsuit, which would not be resolved until December 1993, when an arbitrator issued an award in SPC's favor and ordered Spinnaker to pay $1.4 million in damages. By that time, Spinnaker itself was in the process of merging with two other companies, WordStar International and SoftKey Software Products.
Incidentally, in February 1993, Spinnaker had acquired Power Up Software. Yes, the same Power Up Software that Fred Gibbons had created as part of SPC back in 1983. After spinning off the mail-order catalog division in 1985, the newly independent company initially operated under the name Channelmark before adopting Power Up as its corporate name. Under the leadership of former Software Publishing executive Ed Lauing, Power Up grew to be a $35 million company by 1991. Lauing had thought about taking Power Up public before ultimately opting to sell to Spinnaker. The catalog itself became defunct following Spinnaker's merger with WordStar and Softkey.
Back at Lauing's former employer, 1991 proved to be the year that Software Publishing entered its long day's journey into night. The end of the '91 fiscal year saw record sales of $143.1 million. But that was just a 2 percent increase over 1990. More alarmingly, profits had cratered. After being $19.8 million in the black in 1990, SPC posted a loss of $18.1 million in 1991, the company's first annual loss since it ran out of Janelle Bedke's house back in 1980.
So what happened? Well, the main culprit was SPC's May 1991 acquisition of Precision Software, Limited, a company based in Surrey, England. Precision developed Superbase, a database program for Microsoft Windows 3.0. To put it bluntly, Software Publishing was playing catch-up in the Windows market. The company had started to roll out Windows versions of Harvard Graphics and pfs:Professional Write, but Superbase was already established on the Windows platform. To Fred Gibbons, that made it worth the $25.4 million in cash and stock that SPC paid to acquire the product line.
Gibbons also hoped that Superbase would help jump-start sales of a Windows version of InfoAlliance. What the heck was InfoAlliance? Originally, it was a product for OS/2 machines that made it possible for users to manipulate and update information from multiple different databases on a corporate server. But as OS/2 swiftly lost ground to Windows, there just wasn't a market for InfoAlliance. In fiscal 1991, sales of the product only accounted for about 3 percent of Software Publishing's revenues, while the Harvard Graphics brand now accounted for 80 percent.
On a positive note, Superbase was a widely used Windows database, and its sales helped bolster Software Publishing's revenues to a new record of $156.4 million for fiscal 1992. But net income was a mere $500,000. SPC blamed some of this on the botched Spinnaker deal. The larger problem, however, was that as the DOS software market entered its inevitable decline, companies that relied on high-priced business applications for older PCs faced growing pressure to lower prices. Put another way, the emerging Microsoft Windows application market would not support the high prices that Software Publishing had grown dependent on with products like the now-$600 Harvard Graphics or the $700 Superbase.
Of course, a key source of that downward pricing pressure was Microsoft itself. In 1990, Microsoft released its first version of The Microsoft Office for Windows, an integrated software package including Word, Excel, and PowerPoint. It was the latter that really stung, as PowerPoint was a direct competitor with Harvard Graphics. And as Windows adoption increased, business customers decided it was easier to pay $450 to get three Microsoft programs in a single bundle instead of paying three separate vendors the equivalent amount apiece for their programs.
And while prices were trending down, development costs continued to rise. One reason companies like Software Publishing had resisted developing for Windows was simply the added complexity. During a 1993 Computer Chronicles episode covering that year's European Technology Round-table Exhibition, Fred Gibbons lamented these challenges in Windows development to Stewart Cheifet:
I like to tell people that the first product that we built called pfs:File for the Apple II in 1980 was 30,000 lines of code written by one engineer. One of our current products, Harvard Graphics, [has] plus-or-minus 500,000 lines of code done by a hundred engineers. Well you can see how it's getting harder and harder to build products. So we need some productivity tools in software and I can't be more articulate than that. But I think we'll get to a topic and a subject which will be called "objects" or "components," and we'll snap them together to build bigger systems like they do with hardware.
Attending the Exhibition may have been the highlight of 1993 for Gibbons. It was during that year that SPC started drastically cutting prices on Harvard Graphics and its other remaining products. That September, Gibbons laid off 21 percent of the company's staff, a total of 140 employees. He admitted to the San Francisco Chronicle, "Our business model is broken, it doesn't work. The average price per user is $100 and we have to play the game by those rules or we lose."
The layoffs came as SPC reported its fiscal 1993 results, which showed a 33 percent decline in sales and a staggering loss of $34 million. At this point, all Fred Gibbons felt he could do was keep cutting. In February 1994, SPC announced it would stop further development of Superbase. Three months later, the company sold the product outright to New York-based Computer Concepts Corporation.
This wouldn't be the only acquisition to get undone. In April 1994, Software Publishing Wisconsin, the former Office Solutions, Inc., was shuttered. You'll recall the Madison branch focused on development of SPC's word processors and providing customer support. Product development shifted back to the company's headquarters in California, while the customer support center was sold off to Pennsylvania-based Softmart, Inc.
As part of this downsizing and reorganization, Fred Gibbons also stepped back from his day-to-day management role. In April 1994, he resigned as president and CEO. Gibbons remained chairman, a post he had reclaimed in November 1992 following the retirement of Jack Melchor. But the company would now be in the hands of a new president and CEO, Irfan Salim.
The British-born Salim began his career in the 1970s as a European marketing director for Texas Instruments. He then moved over to Lotus Development Corporation as the general manager of its spreadsheet division. In 1988, Gibbons recruited Salim to join Software Publishing to set up its international division. He quickly moved up the ranks to become president and chief operating officer before Gibbons promoted him to CEO.
The company Salim inherited, however, was not in good shape. After dumping the Superbase line, Software Publishing now depended on Harvard Graphics for 90 percent of its sales. And with Microsoft PowerPoint and other presentation packages now bundled as what essentially free extras with integrated software suites, there was no longer a path to renewed growth focusing on the high-end business market.
That left the low-end market that Fred Gibbons had abandoned back in 1991 when SPC divested its PFS line. Now thanks to dramatically falling PC clone prices and the widespread adoption of Windows 3.1 (and later Windows 95), there was an opening for Software Publishing to come full circle and re-re-invent itself as a producer of inexpensive, easy-to-use software for the non-power user.
Naturally, the way SPC did this was by making yet another acquisition. In April 1995, SPC purchased Digital Paper, Inc., a two-year old startup, for $6.5 million in cash and stock. Digital Paper had created a "digital formatting" technology that automated the layout and design of printed and electronic documents.
At the November 1995 COMDEX show in Las Vegas, SPC announced it would release a "super easy presentation package" for Windows 3.1 users called Type, Click, Present, which was based on this digital formatting technology. The idea was that someone who wasn't a professional graphics designer could create a professional-looking presentation "in a matter of three clicks" by using a set of pre-defined templates. The finished product was released shortly after COMDEX under the revised name ASAP WordPower, which initially retailed for $125. A couple of months later, SPC announced a companion product, ASAP WebShow, a plug-in to view presentations through Netscape Navigator or Microsoft Internet Explorer.
ASAP WordPower generated positive reviews in the press. SPC also touted early group sales of the product to a number of corporate and governmental customers. But perhaps ASAP's real value was making Software Publishing an attractive target for another company looking to make an acquisition of its own. Because that's exactly what happened.
On October 2, 1996, New Jersey-based Allegro New Media, Inc., signed a merger agreement with Software Publishing Corporation. Once the merger closed, SPC became a wholly-owned subsidiary of Allegro. In turn, SPC shareholders received a combined 45 percent stake in Allegro.
Irfan Salim had already departed as Software Publishing's CEO in August 1996, leaving Fred Gibbons to negotiate the final terms with Allegro. Once the merger closed, Gibbons departed as well, moving on to become an investor and board member with a number of companies and later an adjunct professor at Stanford University's engineering school. Meanwhile, a group of SPC vice presidents took positions with Allegro under the new organizational structure headed by chairman and CEO Barry Cinnamon and his chief operating officer, Mark Leininger. In June 1997, Allegro changed its corporate name to Software Publishing Corporation Holdings, Incorporated.
Allegro—now SPCH—was primarily in the interactive CD-ROM business. Prior to the merger, Allegro reported about $9 million in annual sales but like SPC had been running at a loss. Like all business executives, Cinnamon and Leininger thought the solution to their losses was buying other companies. Indeed, six months before the Software Publishing deal, Allegro acquired an English software company, Serif Limited, and its New Hampshire-based subsidiary in exchange for stock. Serif developed a line of desktop publishing and graphics software.
To their credit, Cinnamon and Leininger didn't just buy all these software products to bury them. In February 1998, SPCH announced what it dubbed the "revitalization" of the Harvard brand, teasing a new release of Harvard Graphics 98, which shipped that June. And as the dot-com bubble was now in full inflation mode, SPCH emphasized the new Harvard Graphics' ability to produce Internet-ready websites and presentations.
Of course, the dot-com bubble also meant that SPCH had to try and position itself as a cutting-edge e-commerce company. That prompted more acquisitions and another name change, this time in late 1999 to Vizacom, Inc. Barry Cinnamon had left by this point with Mark Leininger promoted to Vizacom's CEO. This meant that Leininger was left holding the bag when the dot-com bubble went bust.
In April 2000, the Newark Star-Ledger detailed just how bad the bust was for Vizacom. The company's stock, which debuted at its 1995 IPO for $22.50 per share, was now down to just $3. By early 2001, Leininger fell on his sword, and the Vizacom board elevated his second-in-command, Vincent DiSpigno, to captain the sinking ship.
The new CEO wasted no time selling off what few assets Vizacom had left. In February 2001, the managers of the company's Serif division reached an agreement with DiSpigno to essentially regain their independence. Backed by outside investors, the Serif managers agreed to pay Vizacom $2 million over the next two years. As part of the deal, Vizacom also licensed the exclusive rights to Harvard Graphics to the now-independent Serif.
Vizacom would quietly file for bankruptcy in November 2003. Serif, meanwhile, would thrive after the divorce and continued to update and sell Harvard Graphics until 2017. But Serif would become better known for its Affinity line of graphics and design software, which it launched in 2014. In 2024, Australia-based Canva acquired Serif's parent company for $380 million. And just before this podcast was recorded, Canva consolidated the previously standalone Affinity programs into a single application now available under a free-to-use model.
Conclusion
And that’s all for this episode of the Chronicles Revisited Podcast. If you’d like to learn more about the topics discussed in today’s episode, there are links in the show notes. You can also visit my website, Computer Chronicles Revisited, at computerchronicles.blog, which contains full episode recaps and analysis. Thanks for listening, and I’ll talk to you next time.