by Andy Yaschenko
10/22/2003 | 07:20 PM
At first, there was Hawkins. They were three some time later, but Jeff Hawkins was alone at first. Having graduated from the Cornell University in 1979, the newly-baked electronics engineer went to work… Yes, like many other future founders of a company, which would say its word in the industry, Jeff went to work for Intel. He toiled in the research center in Oregon (D1C), then in Boston. Yet, he only stood Intel for three years. Jeff wanted to rock this world, and instead… “I wanted more responsibility. Intel said I needed more subordination.”
So, Jeff turned to a small start-up, established in the Silicon Valley and called GRiD Systems. The company was working on the design of the first portable computer. They didn’t think about pocket computers at that time – simple portability was enough. This was a revolutionary idea for those times as there were not even sub-notebooks then – the year 1982 was running outside. Since his university years, Jeff had a passion for exploring the human mind and wanted to make a step farther. That is, to make a computer directly communicating with the human brain. This was close to an obsession and he lastly followed his wife’s advice to enter the biophysics department in Berkeley.
Three years later Jeff realized he solved the mystery of the human mind. There was a problem, though. No one at Berkeley listened to him. So, in 1988 Hawkins left the university without achieving the longed-for doctor degree. It was unimportant. During his studies of neural networks, Jeff amassed enough research material on the human brain’s image recognition algorithms. Expanding this material further, he formulated an algorithm of image recognition that could be fed into the computer and had it patented. The algorithm was named PalmPrint.
Science was good, but he had to earn his living somewhere. Hawkins went back to GRiD Systems, but for a different role. Licensing his PalmPrint to the company, he became vice-president of research and started getting to the practical implementation of his idea. As for money, they had enough: at about that time GRiD was acquired by Tandy, an electronics manufacturer, which was once known for making one of the first personal computers. Overall, a year and a half after Hawkins’ return, GRiD unveiled one of the world’s first computers with pen input.
And in the meanwhile it was the beginning of a new era – the era of pocket computers. They had even acquired a generic name. John Sculley, Apple’s CEO at that time, christened them as Personal Digital Assistants – PDA (by the way, Apple was into pocket PCs since 1987 when Hawkins was still grinding the sciences at university). The matter was ripe and attracted furtive glances from such giants as IBM, Samsung and NEC.
Meanwhile, Tandy was doing all right, being very active in niche markets and selling GRiDPad to insurance, oil and other industrial companies. But Hawkins, like any other talented man, was also an extremist. He wanted to see his idea in the mass market. He wanted to create a universal and low-cost device suitable for everyday use by everyone. So, Hawkins left Tandy. He left because he felt he would be able to put into life his ideas by himself.
That is how Palm Computing was founded in 1992. Of course, Jeff didn’t have as much money, as, say, Apple spent on Newton. In fact, he didn’t have any money at all. What he did have was reputation and contacts he made when working for Tandy. These proved worth more than money. Hawkins took half a million dollars from each of two venture funds, plus $300 thousand from Tandy as the license fee. With $1.3 million in the pocket, no plan on his hands and no product prototype, he started out.
It’s often hard to single out just one person as the founder of a company. There were Noyce and Moore in Intel, Jobs and Wozniak in Apple. One may be the leader, but the other is indispensable, too. The first hired CEO of Palm was Donna Dubinsky. She was also the second item in the Palm’s success formula. The engineer and the businesswoman were getting along together most successfully.
Donna’s career was first connected with Apple. In 1981, after several unlucky attempts, she got a job with Apple. The company had impressed her with its Apple II. Apple was flourishing then. By 1985, Dubinsky was already responsible for a large share of Apple’s distribution. But there came the critical 1985 year. Wozniak left in February, and in the spring Donna was notified about the changes in Apple’s policy with no place reserved for her division. After unfruitful efforts to stand her ground, Donna was going to follow Wozniak.
However, she stayed with Apple for a while, working under Bill Campbell. She was deep into selling the company’s produce outside the USA. For example, she worked for half a year in Australia. A year later, Campbell, who had become a head of the independent (although sponsored by Apple) software manufacturer Claris, offered her to become vice-president of international sales. In the next four years, working with enthusiasm and gusto, she drove the share of her department to 50% of Claris’ total income. Then Apple bought up the Claris stock and gained complete control over the company. There was no gusto anymore; Donna didn’t want to work for Apple again. She retired, moved to Paris and studied painting for a year…
This year helped her to calm down and get the urge for activity. She called Campbell: “I want to grow up a company," she told him. "I want to be a president. Do you have any ideas?” It turned out that Campbell then headed the most promising, since Newton, start-up in the pen computing field – GO (it was closed up in 1994 after the first failure). Campbell calls up Bruce Dunlevie, “Bruce, I’ve been interviewed for vice-president of technical development for a firm financed by your fund. You didn’t take me, but mentioned that you were looking for CEO for this company. Are you still?”
Well… In fact, Dunlevie and Hawkins had already interviewed many candidates. Anyway, Donna was invited for an interview at Hawkins’ place. Jeff liked Donna’s position, while Donna was all excited about Jeff’s idea. June 15, 1992, she started working at Palm Computing. She later said, “One of my great purposes in life has been to create an environment where Jeff Hawkins could work.” This bunch of a genius technician and a talented manager was typical for those days. Atypically, they managed remain friends until today.
So, they got to work and first decided that the company would only concentrate on developing software for further licensing. They didn’t want to make computers themselves, and thus required powerful hi-tech partners. They did find partners. The Radio Shack trading network, owned by Tandy, was supposed to sell their device. Casio, approached via Tandy, was to assemble the device. Geoworks, Intuit and America Online agreed to port their software for the thing.
What they made was the Zoomer. It appeared in Radio Shack stores, cost about $700 and was powerful enough to compete with PCs (!). It had a tiny keyboard and a somehow working PalmPrint and supported faxes and printers (maybe as an inheritance from the GRiDPad). Anyway, all this, namely high price and lack of user-friendliness, was just the beginning. Yes, because the main idea behind the Zoomer was the fact that it was the first PDA targeted at the mass market. This fact should have concealed all the deficiencies, which might be fixed soon. Fate decreed another way, though.
In August of 1993, two months before the release of the Zoomer, Apple rolled out its Newton, which wasn’t very warmly welcomed. Mostly because of its far-from-perfect character recognition system. Considering that there also appeared other PDAs of the first wave – from GO, Sharp, HP, Toshiba – Palm was in no advantageous position. No one needed just another lame PDA from an unknown company. By the end of 1994, this market had consumed about one billion dollars (about half of them was the money invested by Apple into its Newton project). And none of the companies was a success! This was a total defeat, for Palm as well as for others. Meanwhile, there was a potential demand, and in fact, it was simply colossal. People wanted PDAs, but PDAs didn’t work!
Hawkins and Dubinsky were really lucky to get a third key player into their team. Young Ed Colligan became the marketing manager of Palm Computing in June 1993, that is, the same month that the company announced its Zoomer. Ed could literally sell anything to anybody, but he didn't cope with Zoomer sales. It was hardly his fault. Well, no one really blamed him. Hawkins remarked, “It was the slowest computer ever made by a man. It was too big and too expensive. We did a bad job.”
Fortunately (thanks to Dubinsky!) Palm Computing had just enough money for a second try. Having read the questionnaires of Zoomer buyers, Hawkins was stunned: nearly all of them were PC owners and they didn't want any replacement of it. More than a half of people who bought a Zoomer did it because of the separately (!) sold utility for connecting it to the PC. They didn't need another PC! They wanted an addition, an enhancement to their PCs.
That was a shock. Colligan recalled later: “Jeff basically went off on his own. When he came back, he was like Moses holding the tablets.” The tablets contained only three commandments. The first read: “Simplify the requirements to the recognition algorithm by transferring a part of the work to the user.” Instead of slow recognition of anything man can write, they should go over to fast recognition of special simplified hieroglyphics – Graffiti. The idea of making the user learn a new alphabet for communication with the device was questionable. It met a hard welcome in the team, but was eventually accepted.
The second tablet contained a much more evident thing: the device should be small enough to fit into the breast pocket. The example of the Newton, which was much blamed for its size, was striking. Jeff would make a wooden model of the PDA and go with it everywhere. When he received a call, he took the wooden brick out of the pocket trying to figure out how he would enter data into it. Everybody around were losing their sanity trying to understand why Jeff was playing with a piece of wood. Donna made a correction: Jeff didn’t play, he was inventing.
The third tablet…had the sketch of a cradle with a quick synchronization button. They decided to listen to the users for a change.
Overall, in the August of 1994, three months after Hawkins had sunk into his creative bout, Palm Computing had the vision of a new device. It fitted into the breast pocket, was powered from four AAA batteries, had four integrated applications (a calendar, address book, scheduler and notepad). And it was supposed to cost less than $300. Ed Colligan said: “I remember us sitting around the table when the device was introduced. We all felt tingles down our spines.” The codename of the tingle generator was Touchdown.
The company was toiling on this project for two years watching the market situation getting worse and worse. When Dubinsky and Hawkins came to Dunlevie with their plans, none of their old partners wanted to work with them any more. No one wanted to make a computer to run new software. Once again, Palm Computing was first purely a software developer. The financier listened to the couple and remarked, “'Stop complaining about your partners being unable to do what you want them to. Do you know how to do it right? Then go back to the drawing board and do it yourselves!”
Well, it was easier said than done. Where could they take the money from? Transformation of a software firm into a hardware one meant buying their own fabs and stores, establishment of a distribution channel and so on. Still, there was a roundabout: make contacts with a number of “hardware” companies and create a “fabless” one. That’s what Dubinsky got busy with: sums of money, but mostly the company’s stock, promises of future dividends, rented equipment. They used to joke then in the office that if Donna could rent a cup of coffee, she would surely do it.
These measures solved the production problem, but didn’t help in any way with the sales. According to Donna’s estimates, they needed about $5 million to promote the yet unnamed Touchdown into the market. Considering the market situation, it was close to impossible to raise the money.
The list of potential partners was getting ever shorter until it came to a new name. It was the name of a quite successful company, which Donna had earlier wanted to engage into production of the modem for Touchdown device. The company had boosted its annual sales volumes from $50 to about $900 million in five years. It was U.S. Robotics. The company with headquarters in Illinois wanted very much to move into the Silicon Valley.
Without any serious hopes for any success, Donna called up her friend banker who had worked on USR's initial public offering. He arranged her meeting with John Zakin, vice-president of U.S. Robotics. It was simple then: Zakin saw the Touchdown and was conquered. He promised the necessary $5 million. A couple of weeks later the two parties met in Palo Alto. Zakin was all positive, but never mentioned the money. Dubinsky started to get worried, when Zakin made a sudden offer. U.S. Robotics wanted to buy Palm. The heads of the company wanted to promote their own product and had enough production facilities. It was impossible to resist. $44 million in the U.S. Robotics stock and Palm Computing becomes a division of the modem maker. Touchdown becomes the famous Pilot, the founder of the PDA era, as we know it today.
What do you do, if you have an excellent idea implemented in a single prototype, but lack money to make copies of this prototype to offer to the masses? Seek for money. But what should you do if the market you are going to conquer has already consumed over a billion dollars, and without any real feedback, and no one seems very eager to offer you the necessary sum? Well, there still remains one option. You can sell yourself to someone bigger and more powerful than you (if there is anyone willing to buy). The money and the reaffirmed position will allow you to make what you think fit.
In 1995, when Donna Dubinsky was on her quest for investors to promote the Touchdown device created by Hawkins, she always heard the same thing, “No, thanks.” And then the list of potential investors came to the modem maker, U.S. Robotics. The company was in the spotlight then. Dubinsky wanted $5 million to start selling the device, and John Zakin, U.S. Robotics’ vice-president, was so impressed with the PDA prototype that he accepted all her offers. The impression was even stronger than Donna had anticipated. In a couple of weeks, at their next meeting, Zakin voiced a sudden proposal – U.S. Robotics wanted to buy Palm Computing.
The proposal had no alternative. Of course, the folks at Palm were not too eager to merge with a corporation, to become a gear in the gigantic clockwork, but no one else was offering the so badly needed money. Moreover, Palm Computer was promised a kind of independence in their actions. It was a deal. In the September of 1995, U.S. Robotics purchased a potentially successful product for $44 million and took up its creators along the way. There were 28 people from Palm Computer that continued to work for USR. This was more than enough considering that the OS was developed in collaboration with one of the technological leaders of that time, Geoworks, while U.S. Robotics itself was responsible for the hardware production.
In fact, they were going to start out with small amounts. According to Dubinsky’s estimates, it would be not bad to sell about 100 thousand items in the first sales year. But to err is human. In February 1996, they officially introduced the Touchdown to the public under the trademark of PalmPilot 1000.
The sales actually started in April and first three months were quite calm. But when the market digested the new thing… By the end of 1996, they had already sold 350 thousand devices! The first year and a half brought about 1.5 million units sales volume. They kept on the tempo. By the middle of 1999, there were 4 million pieces sold. By the beginning of 2000 – 6 million. Hawkins’ creation, much to the surprise of its father, turned to be the best selling computer in the history of the industry.
Well, the middle of 1999, the beginning of 2000… By that time, Hawkins and Dubinsky had nothing to do with Palm. Back in 1997, U.S. Robotics itself was bought– by the telecommunications giant called 3Com for $8.5 billion. Of course, that was quite another scale. While Palm felt quite independent inside U.S. Robotics, it became a small, although important, division in 3Com. Two years after the acquisition, in 1999, Palm accounted for 10% of 3Com’s sales volumes – about $570 million. Not bad for a secondary division of a company that is deep in the communication equipment manufacture. Well, 3Com bought U.S. Robotics for its main business, with Palm being a kind of makeweight (The laugh of fate: U.S. Robotics itself proved to be an unnecessary purchase and was set free soon).
Well, the heads of U.S.Robotics/3Com had always been irrationally driven to Palm. After the deal between 3Com and USR, Dubinsky and Hawkins came to 3Com’s head, Eric Benhamou, to ask for their division to be made independent again. They didn’t expect they would get a “No”. However, they were told that Palm would never be separated, as it was too important for the company. They had to succumb and keep on working in the corporate framework of 3Com.
Eric Benhamou
The fruit of their effort was revealed on March 9, 1998, when 3Com introduced a new PDA model, which was to be the next milestone in the industry – Palm III. The device had 2MB of RAM (compared to 1MB in PalmPilot Professional), supported infrared connections, boasted a much improved character recognition algorithm and a slightly different, curvy, design. By the way, present-day PDAs with the pen interface, including PocketPC devices, more or less resemble Palm III in terms of their design.
Of course, there were other numerous innovations and improvements. Quantity transformed into quality: more fonts, a handy cradle design for synchronization with the PC, a new stylus, a folded lid that protected the screen, which had been left intact since PalmPilot and so on. Actually, none of this, including even the new version of the operating system (PalmOS 3.0), went far from simple error correction. They had been just polishing off to the ideal state the successful product, PalmPilot had been.
This could only push the sales volumes ever higher, of course, considering the absence of competition. WinCE devices were only making their first steps in the field and were hard to call successful. Psion had nothing new to offer compared with Palm. However, they had to do something new to keep their market share; simple evolution was not enough. That’s where Hawkins and 3Com collided.
First of all, Jeff and Donna had simply had enough of 3Com’s corporate culture. As many talented people, they preferred to run their own, even small, business rather than be a gear in the gigantic mechanism of the multibillion monster. Secondly, the heads of 3Com were pushing Palm into their main business – communications. Hawkins had other plans for his creation (from today’s point of view, he was wrong, though). Thus came the fateful 1999 year.
At the beginning of it, the company was at the top, if not higher. Palm had about 400 employees (four years before they had only 28 men on staff). Sales volumes were at the maximum: they sold the 3-millionth Palm. They had about two thirds of the PDA market (including keyboard-, stylus-interfaced ones and others – 7.4 million units in 1998 only). Thus, Palm had the first place in this market, ousting Sharp intended for WinCE. Palm licensed its copyrights to IBM with its Palm clone called “WorkPad PC companion”, to Qualcomm with its pdQ Smartphone prototype, to Sun that had licensed HotSync technology. Over 10 thousand developers from small and big companies were writing software for Palm.
That’s how things stood in the beginning of the year that happened to be a revolutionary one for the company. And in the first place it affected the product line, which was split in a very serious way. On the one hand, there continued a slow evolution of existing products like Palm IIIe (the same, but cheaper Palm III) or Palm IIIx (an attempt to do something about the memory. It came with 2MB and with an expansion slot for 2MB more).
On the other hand, they tried to revive these somewhat out-dated products. The result was Palm V, released in the beginning of the year. It was a flat miniature device with a significantly improved screen, although still capable of displaying only four shades of gray. As for bad news, the memory amount remained at the same, 2MB, level (although the Vx model soon fixed this problem). The cradle changed once again (as it did during the shift from PalmPilot to Palm III) and it was not compatible with the Palm III any more. The real good news was the transition from AAA batteries to lithium-ion rechargeable accumulator-batteries that would charge up when the PDA was installed into the cradle.
As for the difference between the two versions of PalmOS – 3.0 and 3.1 – it was quite negligible. Well, the user was now able to adjust the screen contrast. There was a program for training in graffiti and there were better icons… Recalling the past, the difference between Windows 3.0 and 3.1 was much more evident. The new processor had a doubled speed – 33MHz against 16MHz, which was fine already. However, the available programs didn’t run much faster – there were no programs that really required such a speed.
There was also a third direction, such as the implementation of 3Com’s vision – Palm VII. It was actually the same Palm III with an integrated radio module that allowed to seep in information via special sluices. This solution had two drawbacks. Firstly, information could only be received from the sluice, in the coverage area. Thus, non-USA users couldn’t use this feature at all, while in the USA there was no coverage for the entire country. Secondly, the small screen and narrow bandwidth imposed their restraints. Palm VII worked well with e-mail and some other Web services, but was not designed for efficient Web surfing. You could only visit sites from the Palm.net network, which were adapted (sometimes, very severely) versions of fully-fledged sites.
Costing $599 and having such limitations, the device didn’t become a sales hit. Palm.net turned out even less popular. Running a little ahead, I can mention a few numbers. Although the fee for using the network was soon reduced to $45 and the price for Palm VII – to $199 by May of 2000 (and people who bought it with registration at Palm.net paid only $99!), there were only 300 thousand units of Palm VII and Palm VIIx (8MB memory) sold by the beginning of 2001. And the number of Palm.net subscribers was even smaller – 190 thousand.
But that was later. For now, Donna and Jeff lost all what was left of their patience. Having released three new families of Palms, they approached Eric Benhamou once again in July. This time they came to announce their departure. They didn’t want high salaries or options – they needed freedom. At the same time, they didn’t want to wage a war against Palm. They came to Benhamou offering cooperation, with them being outside 3Com. First of all, they meant licensing the OS. Otherwise, Donna & Jeff might license it to Pocket PC or someone else. 3Com would have been the first to suffer from this move.
The company was actually on the crest of the wave, so you shouldn’t think that the guys were leaving a drowning ship. Jeff just wanted to create different devices and Donna… She just wanted to head small companies. As the third member of the team, Colligan (we’ll talk about him later) said: “They didn’t leave because the situation was bad. They left because they could, not because they had to”. Benhamou unwillingly accepted the offer – he couldn’t do the other way. No one knows whether they talked about separating Palm as an independent division…
So, in July Jeff and Donna left the company already knowing what they were going to do in their JD Technologies (they called the new company by the first letters of their names). Hawkins said from the start that they would license PalmOS and would make cheaper PDAs than Palm. After a short pause, in the middle of September 1999, Handspring (they had already changed the name) unveiled its first product – Visor. Who was responsible for marketing? Of course, Colligan, who had left 3Com for Handspring, thus reuniting the brilliant trio of the PDA world.
The second peculiarity of the moment happened a week before the announcement of the Visor, when Jeff and Donna came to show the prototype to Eric Benhamou. He had something to surprise them with. A day before the announcement of the Visor, 3Com was going to announce that Palm was to be made an independent division. I don’t know what Jeff and Donna said or thought on the occasion. However, some time later Hawkins said, “If they had separated us then, when we asked for it, I'd still be working there.”
Why did 3Com change its position? The company had just run out of money. From December 1998 to September 1999, the 3Com stock fell 40% down: stale sales, falling prices for modems and communication equipment. And here they have an opportunity of making public a division costing about $5-10 billion and selling 20% of stock to investors. Extra one or two billion dollars wouldn’t be bad. Moreover, the company changed its policy and was selling subsidiaries to make itself as mobile as before.
So, in March, 2000, Palm Computing went public. PDAs were the hot talk of the day and PalmOS-devices accounted for 4/5 of the market. It was the third biggest public offer of a technological company in history. Here is how the companies ranked after the closing of the exchange on the day Palm was first quoted:
Of course, this was mostly due to the rush of the investors for the new toy, but anyway that was quite characteristic. As a result, Palm acquired $874 million, while 3Com remained an owner of 94.8% of its stock. The rest of the stock was sold to such companies as Motorola, America Online, Nokia.
By the way, who headed this Palm invasion? Dubinsky’s CEO chair was taken by Janis Roberts, then by Robin Adams. He was followed by Alan Kessler who came from 3Com and then – Carl Yankowski.
Carl Yankowski
He headed the company just before it went public, in December 1999. He came from outside – having served as president of Sony Electronics and CEO of Reebok Unlimited. He probably was a good tactic as he managed to retain and enrich Hawkins’ heritage, but proved to be a less successful strategist. He somehow failed such goals as the development of wireless products, invasion into the corporate market, and improvement of the operating system. By the way, Microsoft caught the moment and in the same year 2000 released the new version of its OS – WinCE 3.0.
So, that was the end of the Palm’s golden age. There will be a lot of interesting events in the future, but they will be different. We will watch the leaders – Palm and Handspring – slowly rolling downhill. Still, with a few gleams of sunlight.
So, Palm stepped into the Happy New Year 2001 as quite another company – without its founding fathers (and a mother), but with a new CEO, Carl Yankowski. It was an independent company, separated from 3Com, although the latter still owned the control stock. And this new company faced new problems.
The most urgent one was, of course, the fierce market competition. We shouldn’t forget that in the beginning of this year Hawkins and Dubinsky founded Handspring, which showed fine development dynamics. They were winning the PDA market, ousting the leader, Palm, little by little. On the other side, Microsoft was advancing with its Pocket PC operating system and third-party PDAs running it.
The year 2001 was hard on the computer industry at large and is now remembered for its global market recession with very low demand (although the PDA market suffered less than the PC one). Together with harder competition, this produced obvious results: at the beginning of the year, Pocket PC and Handspring advanced wielding the weapon of price-cutting, while Palm retreated, reacting in the same manner.
Here are the numbers: Handspring approached the year 2000 with sales volumes of $16 million per quarter, and the year 2001 with $115 million per quarter. However, this didn’t save the company from losses, which grew to $7 million in the last quarter of 2000. Of course, the volumes were incomparable: Palm cost $12 billion then (remember $53 billion on the initial public offering day?) against Handspring’s $3.6 billion. The sales volumes differed even more: Palm’s 1.5 billion in 2000 against Handspring’s 270 million. As for the Pocket PC platform, it is hard to cite any numbers as there are much more players in this field. But if Palm’s share of the PDA market was estimated to 60% then and Handspring’s – to 28%, Pocket PC had about 10% (considering the existence of other platforms, too).
The last number isn’t that small at all. Moreover, in the spring of that year the Pocket PC camp was celebrating their jubilee: the millionth sold computer running this operating system. And they did it in a really short time, because only a year has passed since the launch of the first Pocket PC version. Add also that Pocket PC was very active in the corporate market sector from the start, which had always been a dainty bit in terms of pricing. This platform had a higher share there than in the PDA market at large.
Quite naturally, the PalmOS camp was looking for ways to unlock the corporations’ doors. Handspring signed a deal with Ingram Micro to use the wide distribution channel of the latter. Palm was trying to buy Extended Systems. That company was helping corporations to provide the workers with an access to data via wireless devices – this looked like a great opportunity for promoting future Palms with wireless capabilities. The deal has never been signed, though. They intended to exchange the Palm stock for the Extended one, but the former slumped down seriously because of depressing results of the previous quarter.
Well, there couldn’t have been another way. Palm was all in a fever, ending the first quarter with $471 million sales volume and $1.9 million net loss. Meanwhile, a year before the situation was quite contrary: smaller sales volumes ($272 million), but $11 million profit. This tells about both: the profit from an average PDA and Palm’s manufacturing costs. The costs were all-important, as the company actually sold two times more PDAs – 2.1 million against 1 million units. Thus, the average sales price dropped by 20%, of course, but this is not a reason for the losses, especially considering the doubled turnover…
Of course, the company was fighting the negative situation. One solution was quite simple. They would launch new models, more expensive and elite, that would push the average selling price (ASP) higher.
The second trouble, being the consequence of improper company management, was much harder to overcome. Firstly, Palm had seriously overestimated demand and didn’t react fast to the disappointing reality. As a result, the distribution channel got clogged up with older PDAs with all the ensuing consequences: new models should be ushered into the market, but the sellers had to sell out their stock (in spite of low demand!) before doing this. Not only ordinary users were cutting their spending, but also those much-hoped-for corporations. This issue could be solved in a clear, although unpleasant way: by introducing abrupt price cuts and similar measures could work for stimulating sales volume growth, too.
Secondly, although the company was the leader in the field, they definitely overestimated their own potential and lived far beyond their means. They had started up a lot of projects, which would only bring some revenues in a distant future. This issue had its solution, too, and again it was unpleasant: staff layoffs.
That’s what Palm was doing throughout the spring. There were other, external factors, too. For example, the announcement of the new top-end model, m505, had to be made in a hurry when the company was not at all ready for mass shipments of it. This was a reaction to Handspring’s announcement of its flat-screen Visor Edge, which directly threatened the sales of the Palm’s workhorse aka Vx, which was actually bringing in most money. Anyway, the spring model line from Palm was really impressive: the top-end m500, m505 and m515, the improved low-end m105 model and the mainstream m130…
At last, Palm found a modern alternative to the good old Vx, and with long-awaited expansion capabilities, into the bargain. This last feature even allowed competition with Handspring in the marketing field, because new models supported Secure Digital flash cards. Of course, this was not the only innovation. There appeared a new version of operating system (PalmOS 4.0), new batteries with longer life and so on. Top-end models even had a color display!
This meant reaching the next milestone and it was more than enough to answer to the Edge. The m105 model looked appealing, too (yeah, they were low-end, but more adequate than m100). The m130 model was one of the cheapest color-display PDA of that time (I don’t take into account the out-dated representatives of the Palm III family here).
Visor Edge
Of course, a single Visor Edge couldn’t compete with all the rivals, and Handspring reacted with a dramatic price cut throughout the spring. Palm, as I’ve mentioned already, had nothing to do, but follow the suit. For example, the m100 soon went down from the initial $400 to $330. I should also mention that the competitors tried to smooth down the effect of the price wars. Anyway, Palm was the winner of this round. By announcing the new models (the m505 became a new market hit) and cleaning up the distribution channel, the company regained up to 70% of the PDA market by the end of the spring. Handspring’s share dropped to 15%, though it was about 28% in the beginning of the spring.
Nevertheless, this couldn’t tell immediately on the company’s well-being. Other measures – like laying off about 10-15% of the staff (1500 full-time and 400 part-time employees) – couldn’t do it either. The effect would show quickly enough, but not immediately. And the current situation was most depressing, considering how rapidly the company was spending its not so big financial reserves. One of the analysts pointed out something like, “I’ve never seen anyone to go down from an excellent to such a depressing condition in six months only.” That was very characteristic. Moreover, there was the “dead” summer season right ahead.
One of the possible solutions was rather traditional for such situations: splitting the company into two independent divisions. One would produce the PDAs and the other would develop and license (to the first as well as to other market players) the operating system. This solution worked a number of times, but also proved bad in certain situations when more efficient clone makers simply buried the main business.
Having considered all pros and contras, Palm became resolute: in early summer, they announced that the company would separate their software division. One of Palm’s executives, ex-AT&T man, David Nagel, was responsible for this process. He had been supporting this alternative of Palm’s future for months. It must have also been his idea for the software Palm division to be enforced with a purchase of a well-known company aka Be. Be was a creation of one of Apple’s executives, Jean-Louis Gassee, who wanted to develop an operating system sharpened at audio and video applications. One time it had been even considered as the candidate for the new Mac OS.
Jean-Louis Gassee
The company didn’t succeed. The OS was truly an advanced one, but didn’t find its place in the market. Thus, all the technical potential of the developers team was handed over to Palm for $11 million only. Earlier, Apple had offered to buy this company for $125 million, but the owners of Be found it too little a sum. As a result, Steve Jobs’ Next was chosen for the next Mac OS. It was a lucky draw for Palm, really.
It’s sometimes easier to change everything than just something, so the company also revealed its intentions to replace Motorola’s DragonBall processors with much more up-to-date and high-performance processors used in Pocket PC by competitors: those based on the ARM core and built by Intel, Texas Instruments or the same Motorola.
At the same Alan Kessler left the company, too. He supervised the development of the operating system in Palm and headed the Platform Solutions group, which was later made independent. He showed up in a couple of months as the head of the Intransa startup, with 3Com and Palm’s chairman, Eric Benhamou (what a coincidence!), in the administration. Well, this might have been for the better. The situation in Palm indicated that it wouldn’t be worse even if they changed the entire management, not only the head of the software department.
As for the president and CEO of the newly-founded software department, they appointed David Nagel, and it was good. This man should know something about his job as he used to be a senior vice-president at Apple…
David Nagel
Well, I guess it’s the right time to make a short diversion and describe the role of Apple in Palm. I even leave aside the Newton project, the coinage of the word PDA itself… Just remember where Donna Dubinsky came from, who created Be and so on and so forth… Many top officers of Apple later worked in Palm. I can mention Dug Solomon, Palm’s strategy officer, and Satjiv Chahil, the company marketing director. To make the picture complete, I can even cite Steve Jobs himself, who said in the same 2000 that he’d better have bought Palm when he returned to Apple in 1997.
…So, Nagel, ex-senior vice-president in Apple, responsible for R&D, and particularly for Copland OS, AT&T’s chief technology officer, and president of AT&T Labs (the doctor degree in psychology is also not quite irrelevant here) suited well for his role. Once again, about Apple in Palm. Whom do you think Nagel took as a CTO? Of course, the ex-head of the Newton project, Be’s chief operating officer, Steve Sakoman. Quite a mafia family, isn’t it? As for the second division, Solutions Group, which would be producing and selling PDAs, the ex vice-president of Gateway, Todd Bradley, was appointed to head it.
The second opportunity for Palm and for other PDA makers to live on was by expanding the wireless capabilities of these devices. This was clear long before: I could name Palm VII and VisorPhone. But it is really true that wireless communication is more than natural for PDAs. Work in this direction was underway (Handspring even hired the ex-head of the FCC, William Kennard). That summer the first results of this hard work appeared.
The FCC allowed both Handspring and Palm to produce devices capable of operating in the mobile frequency range. Both companies announced their plans (known long before, actually) and Handspring went about by handing a VideoPhone freely to any Visor buyer that would subscribe for any tariff plan at the company website. Moreover, the company promised to roll out new devices with integrated wireless communication before the end of the year. Palm, in fact, also promised the same thing, although unofficially.
It was clear for anyone that this was the future. Donna was quite right in saying that the future of personal computing was also the future of personal communication. But let’s get back to the topic. So, there was summer, the dead season, and the world PDA sales were close to naught, dropping by 21% in just one quarter. The Pocket PC platform was also pushing up (for example, Compaq shipped about half a million iPaqs in the second quarter). So, it was no wonder Palm was losing its finances at the same ratio.
It was rumored on every corner that Palm was going to be sold. There were a number of bidders named, most logical one of the whole bunch seemed to be Sony. The companies were still pulling in their belts. Handspring and Palm both announced more staff layoffs (Palm – about 16% and Handspring – 9%). Palm even gave up the idea of building new headquarters and sold the land. Of course, they both were struggling to squeeze into the corporate market, which seemed like a calm lagoon in the middle of a storming see. However, even though both: Palm and Handspring had a few nice deals in this sector, the fall was the time of Pocket PC 2002, the new operating system, initially targeted at this corporate market.
That’s why there was nothing to get hopeful about. Throughout the rest of the year the companies went on laying off staff, dropping prices, trying to stimulate the decreasing sales and raising capital (quite successfully, though) to live on. It was an Armageddon, really, with money flushing away in tens of millions per quarter. Among new products, I can only recall a quite successful mainstream m125 model (with an SD slot) from Palm and successfully forgotten Visor Neo and Visor Pro that aspired for the same market sector. Handspring announced the Treo family and that was quite another story. They were communicators combining PDAs capabilities, Internet access and voice communication as a regular cell phone. However, although these were announced, they had never been released, so they couldn’t change anything for Handspring.
This way, tired and beaten, both companies dragged to the end of this murdering year. Still, Hawkins and Dubinsky were better off. As for Palm… As I have said above, the company started the year with a doubled sales volume compared to the same quarter of the previous year. The company ended this year with sales volumes 44% lower than previous year and amounting to $290 million only, as if there had been no last two years. The quarter was again unprofitable: they lost $25 million.
Reporting the quarterly results, the company’s chairman of board of directors, Eric Benhamou, admitted the management mistakes made during the past year and promised a change of the situation. About 70% of the company’s managers were to be new people. Carl Yankowski was fired first. He ruled the company into the open market, but didn’t cope with the problems of it.
Of course, Yankowski’s leaving couldn’t do miracles to Palm. Moreover, it was clear that the bad times hadn’t yet ended. At least, the story with the long-promised, but never implemented Bluetooth, continued. The only thing Palm could do was issuing a new Bluetooth SDK for software developers.
As a culmination, Palm’s market share was going ever lower. During the year 2001, its presence in the USA market diminished from 71% to 58%, which was only slightly compensated by the growth of Handspring’s (from 14% to 15%) and Sony’s share (from below 1% to 6%). Thus, the Palm camp was more or less competitive against the common enemy – Pocket PC had 26% against their total 73%. On the other hand, Pocket PC devices only amounted to 15% of the market in 2000.
The Palm camp tried to strike back, more or less successfully. The importance of wireless communication was beyond doubt. That was where they moved to. In the January of 2002, Palm at last released its second device with wireless capabilities – i705 (they had earlier promised to launch it in 2001). The device was something like the m505 with a radio unit that could connect to the Palm.net service for fetching e-mail and doing other things.Considering the “accessibility” of the service to the users, it was hard to expect that i705 would become popular on the world scale. Lack of an integrated keyboard also made work with long messages unhandy (the i750 was supposed to make it easy to send and receive e-mail, wasn’t it?) Handspring understood the last fact quickly enough when they started doing communicators and switched from Graffiti to the keyboard.
Nevertheless, the i705 began selling well enough – about 13 thousand units in the first week. Surprisingly too, most of the buyers also subscribed for the most expensive variant of the Palm.net service that allowed unlimited work with e-mail. It seemed like Palm had managed to interest corporate customers. Moreover, the company also released an application that provided secure wireless connections to SQL-databases via the PDA.
However, the main competitor, Handspring, looked more appealing with its Treo. The device started selling in Asia first, and then, after a short while, came to the USA. The company just once again experienced a severe components shortage that is why they couldn’t assemble enough Treos. Still, this was unimportant as long as the first day of Treo sales in the USA brought about a growth of the Handsping stock by a third. The Treo turned to be a lucky mixture of a PDA with a cell phone, which Hawkins wanted to implement back in the VisorPhone. But this was quite another level of technology. That’s why Handspring also announced that they were going away from ordinary PDAs production.
Meanwhile, Palm, or PalmSource to be more precise, did one more long-promised thing. It held a developer conference. The company showed there a prototype of PalmOS 5, targeted, as the times required, at wireless communication, access to corporate networks, security and similar things. It meant that Palm was catching up with PocketPC 2002, which had officially launched earlier and moved in the same direction.
Software was good, but Palm never forgot its main source of income, which was hardware. The company at last issued a couple of PDAs of the top-end and mainstream class – m515 and m130. They were a good replacement for the m505, with 16MB of RAM onboard, and a cheap color-display PDA for those who were going to buy their first handhelds. There was a nasty scandal with the latter, though. It turned out that m130 had a display supporting 12-bit colors, although the advertisements had promised 16-bit color support. So, there were only four thousand colors per pixel instead of sixty five thousand… At first, Palm just said sorry to the clients who had already bought the m130. When they started suing the company, it presented them a free versions of SimCity.
Meanwhile, Handspring pinned all hopes upon the Treo, and was moving along the only possible course: away from PDAs to communicators. The TreoMail program was released, which provided the most important capability for such a device – e-mail services. In a nutshell, the technology consisted of redirecting the e-mail from the user’s PC to the Treo as soon as the latter is in the coverage area, also when connected to the cell network.
Handspring had entertained good relationships with cell operators. During the first month and a half, there were about a dozen operators supporting and offering the Treo to their subscribers. Among the operators were quite big companies like Sprint. However, this was a move for the future rather than for an immediate effect. The financial results for the previous quarter told that since the beginning of the year, Handspring lost about $23.7 million with sales amounting to $59.7 million only. It was twice as low as in the beginning of 2001! So, it was not a great surprise for the company’s CFO, Bertrand Whitney, to retire in order to “spend more time with his family.”
But again, nothing new can make miracles immediately. The financial results were indicative of the company’s weakness in the PDA market rather than its failure in the communicators market. The situation there was pretty good. The Treo was selling even better than anticipated. For the sales to become mass and tell on the financial health of the company, they needed some time. Meanwhile, Handspring boosted its potential further by launching a couple of new models: the Treo 270, a full-fledged device with 16MB of memory and color display, and the Treo 90, a communicator without … the wireless modem. It had one more peculiarity – an ordinary SD Card slot like in the last Palm models instead of the proprietary slot called Springboard. This made sense – a commonly accepted technology is always better.
But back to Palm. The dramatic reshuffle within the company about half a year ago finally led to the first results. The quarterly financial report mentioned a profit of $2.9 million with sales of $292.7 million. Palm was again on the rise! (Well, it was not exactly so. They listed among the profits the money received from the sold subsidiaries that were previously listed among the losses). By the way, it was the first report of the software division financial results. PalmSource earned $19.5 million with $11.7 million coming from the biggest PalmOS licensee… Palm! In fact, if PalmSource had been fully independent, it would have ended the quarter with about one million dollars loss.
Nevertheless, PalmSource had bright prospects as everyone agreed. For example, in 2001 only, the sales of software for Palm increased by three times, while the company itself was busy writing the fundamentally new version of its operating system, PalmOS 5. It was initially oriented at ARM processors – Palm’s new choice. The invasion into the corporate market was underway, too. They made a client for VPN networks, included Java support, contracted IBM to write a client for access to its e-commerce WebSphere servers…
Besides, they were still preparing for making PalmSource truly independent. This division received an independent board of directors with Eric Benhamou as director, David Nagel as CEO of the division (sounds ambiguous, but that was true), Jean-Louis Gassee, and a couple of others.At the same time, they named the CEO of the hardware division – Todd Bradley, who de-facto already headed it. He immediately faced new problems as the company reported a quarterly loss of $27 million. It was natural as there was no surplus stock at hand. The market reacted with a further stock rate slump. Standard & Poor even excluded Palm from its S&P 500 index.
Well, this was not at all surprising as the PDA sales volumes were still going down. For example, in that quarter they subsided by 18.9%, or down to 1.3 million units. As a result, Palm’s market share became the lowest ever – 42.1%, shrinking by a third throughout the quarter. It was hard to do anything about it, since everybody were in a difficult situation then. Anyway, the curative measures were evident: Palm announced its plans to release in that fall a device for sub-$100 price range, thus trying to trail though the very bottom of the market. It was also active in the corporate sector: signing deals with BMW and medical consulting grand, McKesson…
By the way, Handspring was trying to do about the same thing. Having lost about a half of what they had, the company signed deals with several seller-firms about the Treo promotion. The companies were MarketLink, the government sector-oriented CDW-G, and Global Wireless Data working with cell operators. Handspring also approached one more cell monster, AT&T, and announced the new device, Treo 300, which was an CDMA-version of the Treo 270 intended for the Sprint network.
This is where the good news ends. As for problems,… The company had to stop shipping the Treo 90 and Treo 270 because of defective LCD displays they had. It was a serious blow considering the depressing market situation. As a result, they only sold 170 thousand units in the quarter, which was already good enough for that situation. New staff layoffs followed. Handspring fired 20% of the employees and offered to sublet the offices it had rented before for expansion.
By the way, that was time for Ed Colligan to accept congratulations (or condolences?) for his appointment as Handspring’s president, besides retaining the COO post. He took the chair of Donna Dubinsky, who remained the company’s CEO, though. A funny reshuffle among the three founding fathers in the midst of the crisis, Donna had been actually most responsible for.
Well, PalmSource changed its team, too. They found a CFO, Albert Wood, and enlarged the board of directors by including John Shoven, a professor from Stanford University. The division was doing well. Kyocera and Samsung renewed their PalmOS licenses, Sony Ericsson agreed to collaborate in the Bluetooth field to provide compatibility between PDAs and phones of the two companies. Sony even bought 6% of PalmSource’s stock for $20 million, thus showing its determination to support the Palm platform, and voiced its intention to help in the development of future PalmOS versions.
Thus, it was all normal on the software part, but Palm’s hardware division rose to the occasion, too. They announced two new trademarks, Tungsten and Zire. One was to compete with PocketPC-devices in the high-end sector, while the second (it was that “sub-$100” thing) was to break into a new market for Palm. In fact, the Zire did it during the same fall already. Of course, this was a lightweight version, but let’s be fair. The company has never sold anything for such a price (except special orders). Pocket PC devices have never cost so low, either.
New hopes were in the air. At least, Palm’s share of the market got more or less fixed (41.7%) compared with Handspring’s drop from 22.5 to 6.6%. Considering Sony’s second place with 19.8%, the Palm camp looked most advantageous.
Of course, there were big troubles ahead. The PDA market itself was losing its momentum, even shrinking. The world’s sales volume in 2002 was 9%smaller than in 2001 equaling 12.1 million units in total. And that was not the end. Belts were pulled in once again. Palm fires 19% of the staff, which is about 200 people.
This appeared like a preparation for future battles. First of all, they changed the character recognition system. Xerox was accusing Palm of violating its patents on the handwrite text recognition technology called Unistroke. Actually, Xerox had been doing this since 1997, first suing U.S. Robotics, but now the appeals court in Washington agreed with it.
It meant that all Palm devices could be prohibited one day, all of themthat used the Graffiti technology, violating that patent. Moreover, the big PalmOS licensees (like Sony) would suffer, too. This was too serious a case and too hard to beat. That is why PalmSource licenses the text recognition technology called Jot from the California-based Communication Intelligence. Palm called it Graffiti 2. It was a real laugh of fate: the Graffiti technology was the keystone the original Palm was built on, but it turned to be the weakest chain in the end. Well, Jot had its advantages, too. It is more convenient and allows the user to enter characters in the natural way. Of course, Palm voiced this fact as the reason for the change in the platform.
Still, although Graffiti 2 is user-friendlier and doesn’t make the user learn a new alphabet, this is not the perfect solution. As I have already mentioned, the keyboard is often more convenient. So, Palm licenses the design of the keyboard from the competitors – from Research in Motion, the creators of the popular (in the USA) wireless messenger device BlackBerry (by the way, Handspring also buys the license from RiM, thus ending the patent quarrel with this company). Palm used the keyboard in the direct rival of the Treo – the Tungsten W communicator with an integrated GSM/GPRS modem. That was a nice thing with its color display (320x320 pixels resolution), an SD slot, but also with… PalmOS 4.1 and 33MHz Motorola DragonBall processor! The Tungsten T, released half a year earlier, ran PalmOS 5 and had a 144MHz ARM processor from TI.
PalmSource didn’t sleep, either. It was busy making new licensees. They have already made about fifteen (twice less than Microsoft had for its Pocket PC), and the list of licensees included such monsters as the Chinese Legend. The Chinese market was a dainty bit, never yet tasted by Palm. By the way, Palm also started direct sales there. There were losses, too. HandEra, one of the first Palm licensees, refused to prolong the license and, accordingly, to sell PalmOS-compatible PDAs pointing at the high licensing fee. Well, someone comes, someone goes – Samsung starts selling its first smartphones based on PalnOS 5 – SGH-i500.
They were losing people, too. Steve Sakoman, who came to PalmSource after they acquired Be, returned to his alma-mater, Apple. Maybe he was just tired of the never-ending organizational period. Moreover, having suffered serious losses in the first quarter of 2003, Palm shifted the separation of PalmSource to the coming summer, because there was no money for this procedure at that moment. Meanwhile, PalmSource was already quite ready for this, having brought its first profit in spring.
As for the overall situation… Yes, the Tungsten didn’t take off right. Corporate clients cut their budgets. But the Zire was quick to notch the 1 million sold units mark. As for high-end, Palm launched the Tungsten T2 with twice as much memory (32MB), improved screen with 320x320 resolution, integrated Bluetooth, PalmOS 5.2.1 with Graffiti 2 support and a lot of multimedia software. The summer results were reassuring enough. Although the world’s PDA market is shrinking (by 10% compared to the same quarter in the last year), Palm retains its share of 39.9%. The loss during the last quarter was only worth $15 million.
Handspring didn’t show up, buried deep in its own problems. It thundered in summer. Palm was buying the company, which in fact was its junior brother. Moreover, this junior brother had been often viewed as a smarter and more successful one. When Palm had had to face difficulties, some analysts considered Handspring a potential Palm buyer, although Hawkins never mentioned such plans. It all turned to be otherwise in the end. While PalmPilot was a revelation in its time, and Hawkins was elected a member of National Academy of Engineering, the Treo… The Treo was a good product, but didn’t jump over the bar set in 1997.
Judging by the battle sounds, Donna Dubinsky was not at all pleased with the deal. On the other hand, the new Palm management has made a lot to gain their reputation as quite adequate people and good partners (By the way, Handspring is merging with the purely hardware Platform Solutions division, without any relation to PalmSource, which is to be independent soon).
Jeff Hawkins himself said that the situation has none of the reunion about itself. There’s no nostalgia. Too many things have changed in products and in people. And it seems like there is no place left for Donna in the united company. At least, if Hawkins becomes the CTO of the new company, Colligan heads the wireless solutions division (Smartphone Solutions), she will only sit in the Palm directors board, without any position. Todd Bradley will remain Palm’s CEO, that’s certain.
Currently, Palm is doing not bad at all. It has enough licensees, some of which are big sharks like Sony or Samsung. In a long term, Samsung may quickly grow in the communicators market to the same level as Sony in the PDA market. Palm offers a wide range of models: the Zire is a real sales hit, while the new Tungsten is a worthy competitor to Pocket PC models in terms of capabilities, at least.
Moreover, they have already announced the Tungsten C with integrated Wi-Fi, a 400MHz XScale processor, 64MB of RAM and an integrated keyboard as well as the Treo 600 (whatever name it will have). The new Treo is smaller than previous models, but features an built-in camera and talk time prolonged to 6 hours. PalmOS 6 is also looming in the horizon, which is mostly focused on wireless capabilities, security and such things.
But the main thing Palm has today is the people. It is a kind of “Dream Team” – the best team in the history of the company. It unites the founding fathers, whose creative potential and experience is hard to question, and the new folk who had dragged Palm out of the abyss. We even get the impression that that Bradley, Hawkins, Nagel and Colligan can finally stop defending their positions and attack the Microsoft soon.