As I try to decide if I should write a book, I have been reading or rereading various books about the computer industry.  I just finished Jerry Kaplan’s Startup, written in 1995. Jerry founded GO Corporation in 1987 to develop pen-based computers, and I still think the book has value today for those starting technology-based companies.  It would be hard for me to review this book objectively since I was involved with so many people, situations, and companies it discusses.  The book is richly detailed since Jerry kept many notes and recordings during the more than six years of GO’s existence (1987 to 1994). I wish that I had documented my experiences during my career. It would make writing my book much easier and more accurate!

Reading Jerry’s book brought back many memories of this period for me. Although I don’t think I believed much in those early Personal Digital Assistants (PDAs), I did interact with many of the companies involved, especially GO and General Magic.

Personal Digital Assistants (PDA)

Many companies were formed in the late 1980s and early 1990s around the concept of the PDA, a logical extension of computing’s evolution. First came mainframes, which were not only used by many people at once but were also required a staff to maintain and operate them.  Next came minicomputers, often used only by one person at a time and which did not require staff to keep them operational.  Both mainframes and minicomputers required a fixed location, though sometimes users could be located remotely (time sharing) using a terminal.  With the personal computer, individuals could, for the first time, own and control their own computers.  However, until the development of portable or notebook computers, these, too, had fixed locations: so we went from having computers on floors to computers on desktops.

Even back in the 1970s, then, many dreamed of having a computer that could fit inside a briefcase or, even better, inside a pocket or purse. Today, with smartphones (like the iPhone), we take this entirely for granted.

PDAs required small size, ease of interactivity, and the ability to connect to more traditional, more powerful computer systems.  In a sense, then, these devices were akin to intelligent terminals.  Because size was such a significant consideration, a traditional, typewriter-style keyboard would not fit.  Therefore, it was believed that users would need to interact with such devices using a pen, a stylus, or their fingers.  Several companies developed user interfaces that combined handwriting, drawing, and gestures, including GO, Apple (with the Newton), and Microsoft (with WinPad), to name a few of the more prominent examples. General Magic, wisely realizing that accurate handwriting recognition would not be possible with the available technology, provided a screen keyboard that could be tapped with a stylus. There were also many attempts to add handwriting capabilities to traditional desktop computers.

With the possible exception of Palm, which was formed in 1996, all the companies formed between 1987 and 1997 to exploit this perceived opportunity failed—after raising staggering amounts of money.  The people involved spent years of their lives following a dream they never realized.  Some were and are the most talented people in the computer industry.

Why did so many fail?

Why did so many invest so much money and time into the development of PDAs?  It’s a fascinating question.  Besides the key notion of portability, computers were considered to difficult to use, which restricted the size of the market.  The PDA was meant to be as easy to use as paper but as powerful as a computer; humans would interact with the device through a pen on a small screen. Handwriting was considered a more natural way to interact especially because so few actually knew how to type.

Two things went wrong with this concept.  First, handwriting recognition is a difficult computational problem, and it was just not yet a feasible technology with the portable computing power available at that time.  It might seem amazing that so many great minds did not recognize this, but I have seen this pattern happen time and time again.  Second, people do not really want to use handwriting on their computers, as I think we can now clearly see.  Although we have the computational capability to recognize handwriting on portable devices, almost no one uses handwriting on such devices.  Instead, we are happy to type on glass as proven by today’s smartphones.

So, why did so many venture capitalists and corporations invest so much in PDAs?  Well, I don’t think VCs are very strategic.  I’ve often joked that VCs approach market research by counting the number of entrepreneurs coming into their offices with virtually the same business plan.  If so many smart people want to start companies to do the same thing, they figure that thing must make sense.  Once a few VCs begin to invest in a particular area, other VCs jump in as followers.  Then, VCs need to find others to invest the bulk of the next rounds of financing.  Getting large corporations to invest is especially helpful, as they are generally thought to be less sensitive to valuation.  A few large corporations succumb to the notion that the idea must make sense because so many VCs have invested.  Other large corporations, fearing they’ll be left out in the cold, then decide to start internal projects, which eventually fail.  Seeking an alternative, the corporations buy the startups instead.  If the timing works out right, the VCs and a startup’s employees can earn significant returns before the business fails, but if it starts to fail before it sells, only the early-stage VC investors usually make any money.

Considering the history of the PDA as an example of such institutional behavior, we might ask ourselves if this kind of thing is still going on.  I think it is and would point to Virtual Reality as one example.  Over $3 billion was raised by VR startups in 2017.  The most prominent being Magic Leap who has raised $2.3 billion dollars to date.

GO never got going

Turning back to GO, it turns out that Intel was an investor.  Though I knew that when I started Jerry’s book, I forgot the full story.  As frequent readers of this blog will recall, I was co-founder (along with Les Vadasz, one of the founding team of Intel)  of what became Intel Capital, we called it Corporate Business Development (CBD).  In Jerry’s book, I learned that I had not initiated Intel’s investment in his company.

I started doing early-stage investments and acquisitions for Intel in 1988.  I had a few people working for me, and I reported to Les Vadasz, who managed Intel’s Systems Division in addition to trying to manage me.  Then Les went on sabbatical and taught at Harvard for a while, during which I reported to Frank Gill (who had taken over from Les) and later to Dick Boucher (who kept Frank from firing me—Frank was not interested in venture investing).  When Les returned to Intel in 1991, he decided to devote himself full-time to acquisitions and venture investing, and together we set up CBD.  Before that, some investments were made directly by the various business groups at Intel.

John Doerr, a former Intel employee an partner at Kleiner Perkins, brought Jerry Kaplan to meet with Andy Grove about making an investment in GO.  Jerry’s book describes those discussions.  For Intel, the key was for GO to commit to move their product from the Intel 286 to the Intel 386 processor and to agree to always port GO’s operating system (PenPoint) first to the Intel Architecture.  Though Kaplan was unhappy about these requirements, he needed Intel’s money and blessing to help GO close a big round of financing.

At some point, Les asked me to meet with the GO team, which was having some problems.  I went to their offices in a large building in Foster City, San Mateo.  There I met with the founder, Jerry Kaplan, and members of his team, including the legendary Bill Campbell, recently joined as CEO.  I had met Bill many years earlier when he ran Claris, then the software arm of Apple.  Bill and I later developed a friendship.  He was a fantastic man who sadly, died a few years ago.   After GO failed, Bill went on to a very successful career as CEO of Intuit, board member of Apple, and advisor to many prominent Silicon Valley companies.  He “coached” Steve Jobs, as well as Larry Page and Sergey Brin (the founders of Google).  Bill and I both lived in Palo Alto when I worked at Intel.  We would occasionally meet up there to chat.

I’m not sure whether Intel lost money on its investment.  When GO started, it had a hardware group.  Later, it decided to concentrate on developing its pen-based operating system, spinning out the hardware group into what became EO, which got a large investment from AT&T.  Eventually, GO and EO both merged. AT&T invested enough to become majority owner of the new entity.  Part of that investment went to fund the combined company, but part of it bought shares from existing investors, including Kleiner Perkins and, I assume, Intel.  According to Jerry, Kleiner Perkins got their money back and more, while the founders and employees of GO got nothing in the end.  Kleiner Perkins was always very good at making sure they did well even when their portfolio companies did not, but that is another story.

At the end of the book, Jerry rightly wonders not why GO failed, but why it lasted so long and could raise so much money.

Besides Jerry Kaplan and Bill Campbell, GO had many successful alumni: Stratton Sclavos later founded Verisign ( a company in which Intel invested) Randy Komisar became CEO of LucasArts and later a partner at Kleiner Perkins, and Mike Homer was VP of Marketing at Netscape when it made its very significant IPO.

The magic moment

General Magic, spun off from Apple in 1990, was perhaps the most impactful independent PDA company—impactful, that is, in the sense that it was a highly visible failure.  It was founded by Marc Porat, Andy Hertzfeld, and Bill Atkinson.  Andy and Bill were the two people most critical to the development of the Macintosh.  Marc, strangely enough, produced a film commissioned by Digital Equipment Corporation to celebrate that company’s 25th anniversary, a movie in which I played the “leading role.”  I knew Marc very well professionally and personally.  He was a masterful storyteller which, combined with the pedigree of the other founders, made a very compelling pitch, especially to corporate partners.  General Magic raised $200 million from 16 strategic partners, including Sony and AT&T.  They also raised $96 million in an IPO. The valuation was $834M The stock price doubled on its first day of trading.  The professional investors who were able to buy stock in the IPO sold off their shares and price dropped back to close at the offering price.  That was the company’s high point.  Its PDA product was not at all successful.  Marc Porat was replaced a year after the IPO, and most of the team left.  The company staggered along for another eight years before bankruptcy.

General Magic’s device was more or less an intelligent terminal.  They had a very sophisticated model of networked computing, with various computational tasks assigned to computers across the network.  Though their interface still used a pen, General Magic avoided handwriting recognition, displaying a keyboard on the screen which one could tap using the pen.

I can remember that I was impressed with some of their concepts, especially their networking model which created a distributed system. I am pretty sure that Intel did not invest, but I’m not sure why. I know I had several meetings at General Magic, and I can remember explaining the company to Les and others.

Even though General Magic came out of Apple, Apple only owned 10% of the company.  John Sculley, Apple’s CEO (Steve Jobs had already been forced out), started an internal, competitive PDA project, called Newton.  Many in General Magic felt that Apple stole ideas from their company. Newton had insufficient connectivity and relied on handwriting recognition, which was not very accurate, but it completely sucked the air out of the market for General Magic.  The first Newton units shipped in 1993, and the product was not successful.  The last Newton was made in 1998, and its failure helped turn the market against the idea of PDAs.

The alumni of General Magic are very impressive.  Besides the founders, Ton Fadell later co-invented the iPod and iPhone and later founded Nest.  Megan Smith later became the CTO of the United States under President Obama (a position since eliminated).  Andy Rubin founded Android, Pierre Omidyar founded eBay, and Steve Perlman co-founded of Web TV, among many more.

Recently, a documentary on General Magic was released.  I have not yet seen it, but here is the trailer.

WinPad

Besides Apple with the Newton, many corporations had PDA projects.  WinPad was one of these, which I only mention because I was involved.  In 1992, Andy Grove asked me to take the lead in coordinating with Microsoft to deliver their needs for consumer products.  Intel had no interest at that time in the consumer market, but Microsoft did.  My job was primarily to keep Bill Gates happy and make sure they did not turn to our biggest competitor, AMD.  My counterpart was Rob Glaser at first.  When he left, towards the end of 1992, to found Real Networks, I worked with Craig Mundie, who eventually became Microsoft CTO.

We identified three major consumer projects.  The first project, which never really got started, was a game console.  The second was an intelligent set-top box. WinPad was the third.  I’ve written a lot in the past about the set-top box project (internally nicknamed Pandora’s box at Intel), because it eventually led to the creation of today’s residential broadband network (with no thanks to Microsoft).

I coordinated a team to work with Microsoft to prototype the WinPad and help develop partnerships that could take it to market.  I thought the WinPad had a chance because it used the Microsoft code base and therefore would be appealing to corporations. I did not think it would be a consumer product.  Companies like Compaq, Dell, and even IBM started projects based on the WinPad.  However, the project ended sometime in 1993, probably because it could not meet its cost goal.

Visionaries or Dreamers

I like to say a Visionary is a Dreamer with a PR department.  Much can be learned by studying the history of the PDA.  Billions of dollars were spent pursuing an idea that was not only technically unfeasible but also undesirable, most likely, to its target market.  Brilliant entrepreneurs spent many of their most precious years chasing the dream of the PDA.  Some of the most successful venture capitalists invested hundreds of millions of dollars in early-stage PDA companies, and key firms in the computer and communications industries spent hundreds of millions of dollars either in investments into early-stage companies or on their own internal projects.  PDA companies taken public later failed entirely.

Just as we have examples of such failures, we also have examples of success.  Once early-stage companies like Amazon, Google, and Facebook grew to become enormously successful.

How is it possible to distinguish between the two?  You might have to wait for my book to find out.

History of Pen Computing

Bill Campbell

Bill and Andy’s excellent adventure

More on GM

GM IPO

Case Study of GM