Computer Chronicles Revisited 43 — Trip Hawkins, John Merson, Ben Anixter, and Richard O'Brien

This next episode continued the previous discussion about the noticeable slowdown in the computer industry during the summer of 1985. This time, the focus was on software and hardware manufacturers, including three companies that not only survived the slowdown but remain major players in the industry today.

Like the previous show, Stewart Cheifet did his cold open from a street location in Silicon Valley, specifically Semiconductor Drive in Sunnyvale. He rhetorically asked about the future of the high-tech business given the ongoing slowdown in computer sales.

In the studio, Cheifet and Gary Kildall showed off two of the latest high-profile personal computer releases, the Atari 520ST and Commodore’s Amiga 1000. Cheifet noted both machines were praised for their good technology and low price. Yet industry analysts were saying that neither machine may make it in today’s marketplace. Just how difficult was it to bring out a new product? Kildall said the key issue was price. The Amiga cost $1,500, while the ST cost $800. Both machines came with a Motorola 68000 microprocessor and 512 KB of memory, so they were at least as powerful as the early IBM PCs. The question, however, was how these machines would fare in the commercial market and home market, respectively. The commercial market meant things like word processing, spreadsheets, and graphics that would help these machines compete against IBM’s stronghold. The home market presented a more difficult target. The manufacturers had to come up with software that people really wanted to use in the home. But if Commodore and Atari couldn’t make it with these computers, Kildall quipped, “We might as well hang it up.”

Could the Software Industry Survive on $50 Applications?

We go right into the first of two round tables for the episode. Trip Hawkins, the founder and president of Electronic Arts made his second appearance on the program. He was joined by John C. Merson, the vice president of marketing with Ashton-Tate, the developer of the popular database application dBase III. Kildall opened by asking Hawkins about a recent report suggesting that even as hardware sales had slowed, software sales were up 60 percent. Why the contradiction? Hawkins said the industry had been going through a phase of consolidation and overall 1985 was a slower-growth year than what some people had expected.

Hawkins continued that there were three major reasons for this. First, consumers were “tremendously confused” due to the conflicting promises and news they’d heard over the past few years. He compared it to the record industry back in 1948, when CBS introduced the LP format for the first time and RCA responded with the 45 format. Hawkins said this led to a four-year slump in record sales as consumers waited to see what standard would prevail. Second, there was an ongoing technology war and every year there were a few new computer models that were incompatible with existing systems. Again, he used an analogy: Would you buy a new and improved television set if it didn’t allow you to watch any of the current television shows? This created additional consumer confusion. Kildall asked if Hawkins was referring specifically to the Atari 520ST and Amiga, which were incompatible with the older home machines. Hawkins said yes, that was right. Hawkins continued with his third point, namely that there had been “real chaos” in distribution channels for computers and software.

Cheifet then asked Merson for his view: Was the software industry not in the same situation as the hardware industry? Merson replied that the software industry had to deal with many of the same factors as the hardware industry. There were too many players trying to bring too much product to market–and distributors and customers really could not absorb it all.

At the same time, he added, people tended to buy software throughout the life of a system they owned. As they became more conversant with the system they wanted to do more with it. So 1984 was a fantastic year for hardware, and he thought as those buyers matured, they’d want more and more software.

Kildall, continuing one of his themes from the prior episode, then brought up the question of “price erosion.” He noted that in his own experience going through multi-channel distribution, the percentage taken at each level was fairly high. But as prices at the retail level continued to go down, software developers like him got to the point where they’re saying, “Why should you even be in the business?” (Cheifet interjected that what Kildall called “price erosion” was great for consumers like himself.) Merson said that in many ways, software prices had been moving up and “defying gravity.” He said there were currently software products being sold on the mass market for between $600 and $700, which would have been unheard of just a couple of years ago.

Kildall followed up, asking Merson to clarify what he meant by the “mass market.” Merson replied that he meant the business software market, which covered about 4,000 stores across the United States. He said Ashton-Tate’s software was selling “in large numbers” at the $600-$700 range in those outlets. For example, Ashton-Tate’s Framework had sold over 70,000 copies last year.

Kildall turned back to Hawkins and asked if it would be harder to sell a $600-$700 software package on a machine like the Amiga, where the customer had to pay $1,500 just to buy the machine. Hawkins said it would be. (He also corrected Kildall, noting the Amiga 1000’s list price was $1,295 for the basic configuration without a monitor.) He believed that most of the software sold for the Amiga would be priced under $50.

So how can you make any money at $50, Kildall retorted. Hawkins said it would have to be done on volume. He added that falling prices were in the best interests of the consumer, to the point that it allowed the manufacturers of software and the creative people making programs to continue to be attracted to that business.

Kildall asked what kind of install base would these lower-cost PCs like the Atari and Amiga need to have by the end of 1986 to be sustainable platforms. Hawkins said that much of the software that Electronic Arts sold went to new buyers. That is, many people bought a lot of their software when they bought their first machine. But since distribution was such a jumbled mess right now, it really mattered more which machines the retailers decided to carry as opposed to the installed base. So Electronic Arts didn’t look at the market in terms of installed base.

Kildall continued to press his point, asking Merson that if $50 software became the norm, would we see a $49.95 version of dBase III? Merson said we’d see $49 versions of some business software. Indeed, companies like Borland had already brought desktop accessory software to market at under $50. So it could happen. But Merson added that customers looked at business software not just in terms of what they were getting today, but in terms of making a major investment of time and energy into learning the product. That investment far outweighed the cost of the software package itself. More to the point, customers wanted to be sure that the company that made the software would be around a long time to support it. In that sense, there was more “brand consciousness” among customers of business software.

Cheifet pivoted to a discussion about Ashton-Tate’s recent acquisitions of other software companies. Did Merson see that as an industry trend? Would there be more consolidation in the software market? Merson said he thought so. This was a period of consolidation, there were too many software companies for the size of the market and its projected growth. He added that we’d probably past the point where a one-product company–even a company with one very good product–could be viable on its own. As these companies got together and shared sales, distribution, and marketing infrastructures, the surviving companies would be stronger.

Kildall added that customers would also look for more integration within software products. Merson agreed, noting that customers wanted products “to talk to each other”: What good was the greatest word processor in the world if it couldn’t exchange data with the greatest database in the world? Customers were now more insistent on software compatibility across a broad range of products.

Returning to yet another theme from the prior episode, Kildall asked what software product would finally make the home market grow? It probably wouldn’t be a database program, he quipped. Merson said he thought the three most important home applications were work, work at home, and homework. That is, home computer users wanted to use their computer to work on projects they took home from the office, run a business out of their home, and for children to work on their homework. When there was a significant installed base of PCs in the home, Merson said we’d probably see work-oriented software as a leading category.

Cheifet brought the segment to a close by asking Hawkins if he saw $50 software on the Amiga as applying to business software and not just games. Hawkins said he was talking about all kinds of software, including creativity, productivity, and games. He added that the market for home consumer electronics, including personal computers, were driven by men who were primarily interested in entertainment. So he believed that when consumers realized the entertainment experience was profoundly different on a personal computer, that was when that market would be willing to spend the money on PCs.

Ex-Apple Employees Bullish About New Startup Venture

Wendy Woods presented a brief segment featuring two former Apple Computer employees who had started their own software distributor. Woods opened by noting that Apple itself had been struggling during the downturn, laying off 20 percent of its staff during the summer of 1985. If the “dream” of Apple hadn’t died, it had certainly been curtailed, she observed.

Woods spoke with Dave Larson and Bill Cleary, who left Apple to run a new company called International Solutions. Cleary told Woods they were a lot more “frugal” at their new startup. Back at Apple, for instance, they had spent over $2 million in one day launching the Apple IIc.

Woods described International Solutions’ business as “marketing unique Apple software and hardware from abroad.” They also shipped American products overseas. These products exploited mouse technology to enable the Apple II to emulate a Macintosh-style user interface. Woods added that Larson and Cleary had started this business without any venture capital funding. Larson said he saw the market now experiencing a “resurgence of the entrepreneur,” with a new group of startups successfully launching products and product concepts. Cleary said he was also “bullish” about the market, adding they wouldn’t be working 16-hour days if they thought the business was over. This was just a transition period.

Semiconductors Slumped Due to Excess Inventory

The second and final round table featured Ben Anixter, the vice president of corporate marketing at chip manufacturer AMD; and Richard O’Brien, a corporate economist at Hewlett-Packard. Kildall asked Anixter if there was currently a slump in the semiconductor industry. Anixter replied that there had been a “severe” slump. Beginning in September 1984 there had been a very rapid slowdown in orders and shipments, which had now stabilized, but AMD’s sales were still down about 25 to 30 percent compared to this same time last year.

Kildall asked if competition from Japanese semiconductor manufacturers was the cause of this slump. Anixter said it was one concern but not the only one. There was also a tremendous accumulation of inventory that needed to be worked off plus a lot of “over-optimistic” forecasting, which resulted in a lot of companies in the PC and other industries buying a lot of stuff they didn’t need.

Cheifet asked O’Brien about the effect of the slowdown on Hewlett-Packard, which tended to focus on the scientific market. O’brien said the slowdown had been going on for about nine months at this point. He said that 1984 was a phenomenal year, and if you compared where H-P was today in 1985 compared to 1983, the growth over that two-year period was fairly good. But people’s expectations for 1985 assumed a continuation of 1984, which O’Brien said was more of an anomaly than where we were today.

Kildall asked O’Brien where he saw the “major” decline in the market. O’Brien said H-P broke the industry out into four areas: instruments, computers, communications, and semiconductors. The weakest area by far was semiconductors, he said. The computer area was flat in terms of growth. But the instruments and communications areas continued to show reasonable growth.

How would H-P pick up the computer part, Kidall asked. O’Brien said the market for capital goods overall was flat. So when those customers were financially healthier, then you’d see a recovery in the computer portion of the market.

Cheifet turned back to Anixter and asked about reports of excessive spending at AMD, such as a $100,000 Christmas party and the “elaborate things” that used to go on in the business. Did people used to be “too aggressive” and “too fanciful” in the industry? Anixter said he didn’t think so. (He added that Christmas party actually cost more than $100,000.) That was part of “sharing the wealth” with a very loyal employee base. He added that AMD had a “no-layoff” policy, which would make it easier for AMD to help get back up and running when there was an upturn in the market. Cheifet clarified that AMD did furlough people when business was down. Anixter said AMD used “short work weeks,” so that everyone shared in the bad times as well as the good times.

Kildall observed that as an outsider looking at the chip industry, once a particular type of chip became successful, such as a memory chip, then it “goes across the ocean” and was built over in Japan at a lower cost. That forced American companies to keep innovating. So what was AMD looking for in terms of innovation? What new areas would it get into? Anixter said the whole part of the product line was important. You couldn’t build a computer with just memory chips. You had to have innovative products in other areas to make memory work. He added that memory was only about 20 percent of the semiconductor business. What separated the companies that customers wanted to deal with was the ones who made the innovative products in other areas, such as networking, graphics displays, microprocessors, and peripherals.

Kildall asked if AMD had actually seen growth in demand for products like graphics and networking chips, or was that just speculation? Anixter said it wasn’t speculation. It took a long time to get things going, so AMD was talking to customers about products that wouldn’t be ready for production for another year or so.

Following up, Kildall asked if demand for local area networking specifically would help offset the current decline. Anixter said that while nobody liked a decline in business, the resulting lowering of prices helped stimulate demand for new applications for semiconductor products.

Cheifet asked O’Brien if companies would be more conservative and less willing to take risks because of this decline. O’Brien said he didn’t think that would happen in the long run. People might be a little bit more conservative over the next year. Things weren’t deteriorating further, he said, but they hadn’t started to turn around yet.

To close the segment, Cheifet returned to the subject of Japan, asking if there was a serious long-term risk that the United States would “lose control” of the semiconductor industry. Anixter reiterated that innovation drove the industry, so you could compete if you could innovate. Moving manufacturing offshore certainly helped to compete in the international market. But “brainpower” and innovation would keep America and “our side” in good shape.

Would the Computer Industry Suffer the Fate as Auto Manufacturers?

Paul Schindler offered his first closing commentary of the new season. Admitting that his role was to be a “contrarian,” he said he saw things getting better for the computer industry. He said the “good times” of 1984 were an aberration but so were the “bad times” of 1985. The real long-term growth of the computer and electronics businesses lied somewhere in between these extremes. He said there was nothing “fundamentally different” about this recession in that there would not be a major restructuring like with the automobile industry. Essentially, there would continue to be a multi-vendor computer industry.

Jobs’ Apple Sell-off a Likely Precursor to New Company

Stewart Cheifet presented this episode’s “Random Access,” which featured news headlines from late September 1985.

Ashton-Tate Fell Victim to Single-Product Problem

Three of the four companies represented in this episode–AMD, Hewlett-Packard, and Electronic Arts–are still in business today. The outlier is Ashton-Tate. This may be a forgotten company now, but in 1985 it was considered one of the major publishers of business computer software along with Microsoft and Lotus. So why did it fade into obscurity?

Well, as John Merson himself said in the episode, the era of a software company relying on a single product for its success was coming to a close. And despite aggressive efforts by management to diversify its portfolio, Ashton-Tate would continue to live–and ultimately die–by its signature product, dBASE.

The history of dBASE and Ashton-Tate is somewhat interesting, starting with how both were named. At the time of this episode, Ashton-Tate was riding high off the success of its $700 dBASE III package. As you might expect, dBASE III was the successor to dBASE II, which was released in 1981. But dBASE II was not the successor to dBASE I, which never existed, at least under that name.

Let’s back up a bit. In 1980, George Tate and Hal Lashlee formed Software Plus Inc (SPI). This company did not originally develop or publish software. Rather, it ran a mail-order software distributorship out of a two-bedroom apartment.

One of SPI’s customers asked about a database program called Vulcan. This prompted Tate and Ashlee to meet with Vulcan’s creator, Wayne Ratliff. At the time, Ratliff worked at NASA’s Jet Propulsion Laboratory in Pasadena, California. He developed Vulcan in his spare time to assist him in plotting strategy for a football betting pool. He’d tried selling Vulcan–named for the Star Trek character–through magazine ads but found little success.

At that point, Tate and Lashlee decided to get into the publishing business and market Vulcan themselves. They signed a contract with Ratliff where he granted them exclusive marketing rights in exchange for royalties on any sales. Tate and Lashlee then enlisted the services of a Los Angeles-based advertising executive, Hal Pawluck, to help them market their new product.

It was Pawluck who renamed both Vulcan and SPI. (The name Vulcan was already in use by another commercially available software program.) He decided to call the database software dBASE II, apparently because it would make customers think it was an improvement over the (non-existent) dBASE I. As for the company, Powluck said SPI should use the name “Ashton-Tate” as a publisher. You’ll note that the owners of SPI where Lashlee and Tate, but for some reason Pawluck thought that “Ashton” sounded better than “Lashlee.”

By January 1982, SPI/Ashton-Tate had 17 employees working out of that two-bedroom apartment. dBASE II had proved to be a major hit. The company reported $3.7 million in revenues for its 1982 fiscal year, although it still ended the year with a $313,000 operating loss. Tate, the company’s president, decided it was time to hire an outside manager. David Cole, who had been working as a vice president of marketing at another tech startup, accepted the offer to become SPI’s president after briefly serving as a consultant.

Within a year of Cole taking over, the company’s revenues increased nearly 500 percent. During 1983, Ashton-Tate acquired all of the rights to dBASE II from Ratliff, who also joined the company as a full-time employee. Cole then took Ashton-Tate public in November 1983.

In an effort to diversify, Ashton-Tate launched its second major product, an integrated office suite called Framework, in January 1984. Similar to dBASE II, Ashton-Tate initially licensed Framework from an outside developer. The company also released dBASE III in June 1984. Both programs proved financially successful, although dBASE would remain the company’s single dominant product.

There would also be another significant transition in Ashton-Tate’s management during 1984. Co-founder and chairman George Tate was found dead at his desk in August 1984, having suffered a fatal heart attack at the age of 40. Less than three months later, in November 1984, David Cole unexpectedly resigned as president to join computer magazine publisher Ziff Davis. Cole’s replacement at Ashton-Tate was the company’s head of sales and marketing, Ed Esber, Jr., who previously held a similar role with VisiCalc publisher VisiCorp. Esber would lead Ashton-Tate for the remainder of its time as an independent public company.

Ironically, it was the low-cost competitor mentioned by John Merson, Borland Software Corporation, that would ultimately buy Ashton-Tate. After the follow-up to dBASE III–that would be dBASE IV, released in 1988–proved to be a complete disaster, Esber went looking for a buyer. Borland ended up paying $439 million to acquire Ashton-Tate in September 1991. Borland tried to update dBASE for the Windows era but ultimately abandoned the project in the face of stiff competition from Microsoft Access.

As for John Merson, his own tenure at Ashton-Tate proved remarkably brief. Merson had joined the company in April 1985 as vice president of marketing. He previously held the same title at The Computer Factory, a New York-based computer retail chain. But in November 1985, just a couple months after his Chronicles appearance, California-based Ashton-Tate announced that Merson had resigned to return to the east coast.

Merson is a veteran of the Vietnam war and has since 2015 served as a board member at Children of Vietnam, a nonprofit organization that focuses on reducing child poverty in Vietnam. He also works in real estate development, renovating residential properties in New England.

EA Thrived After Moving Past Hawkins’ Console Reluctance

Trip Hawkins strongly disagreed with John Merson’s view that working from home would drive computer sales in that market. At a December 1985 industry event, Hawkins said, “You’re kidding yourself if you think Joe Six-Pack..is interested in even the slightest way of doing more work when they get home.” Not surprisingly, Hawkins saw people buying computers primarily to play games.

To that end, Hawkins bet heavily on the Amiga 1000. EA spent about $600,000 developing launch titles for the new computer, according to news reports from the time. But as Hawkins himself said when correcting Gary Kildall, a base model Amiga 1000 without a separate monitor retailed for around $1,300 in September 1985. About a month after this episode aired, Nintendo of America began selling the Nintendo Entertainment System (NES)–the redesigned Nintendo Family Computer from Japan–for around $180. Commodore sold about 4 million units of the Amiga 1000 and its successors worldwide between 1985 and 1994, according to Popular Mechanics. In contrast, Nintendo sold over 33 million NES consoles in North America alone.

Yet Hawkins was, to put it charitably, slow to embrace the revived video game console market. As game historian Jimmy Maher explained in a 2018 post for his blog The Digital Antiquarian, Hawkins faced a boardroom revolt–which nearly turned into a “coup”–over the issue. Hawkins was still unwilling to deal with Nintendo due to its restrictive licensing terms, but EA would embrace and thrive on a rival platform, the Sega Genesis, which was in many respects was a souped-up version of the Amiga 1000. Both machines were built around the same Motorola 68000 microprocessor that also powered the Macintosh. But the Genesis was more user-friendly and less expensive than a full-fledged personal computer.

This time, EA’s bet on a new platform paid off. Maher noted that about 25 percent of EA’s sales in 1990 were for Sega’s machine. And in 1991, the year the Genesis “really started to take off” in North America, EA had four of the ten best-selling games on the platform. Among the successful Genesis titles was John Madden Football, a property that would come to define EA as the premiere publisher of sports games to the present day. (EA would later become a publisher for Nintendo’s consoles as well, although this largely happened after Hawkins stepped aside as CEO in 1991.)

Notes from the Random Access File