Letters : The last thing I wanted to do in the story titled HP reprising DEC's fate: Microsoft VAR? was to sentimentalize Digital, or suggest it had a divine right to exist, no matter what mistakes it made.
As Len Dobrowsky points out:- "Don't go deifying Ken Olson. Digital was a great company from its start until about 1972, but then it fell into irreversible decline because of his consensus management policy… the high margins just hid bloat."
But the piece did spark many of you to provide reflections on how Doing The Right Thing is more difficult now than in the Olsen era, and the story was as much the start of a conversation about how finance capital serves technology - or doesn't - as it was about tech companies.
"Computing is in a classic Keynsian demand side crisis, only is Microsoft going to use its billions to give the software away?" asks Hanifi Houbart, who doesn't write to us very often, but when he does, it's pretty good. "The effect in both the US and EU of government policies such as DMCA, EUCD, Hollinger and such, is to create additional decline in demand."
DEC and HP veterans and fund managers have contributed to this. One HP veteran sees parallels with DEC.
I am a grey beard from the "bad old days" of HP (been using them for 23/24 years). In those days, they were proud of their hardware and software (MPE and RTE as they were then) and were actively creating their own flavours of Email, editors, data dictionaries and so on (I'm dealing mainly with the 3000 here). These were excellent bits of software - not cheap but easy and quick to use. Then along came the HP/UX boxes and software. Same hardware, different operating system but the cost of the 3000 hardware was (and still is) around 2.5 to 3 times that of the equivalent HP/UX box. Therefore, for a unit product in the UNIX world, the same box would sell for 2.5 to 3 times in the MPE/iX world. People still bought the 3000 and do today.
Nowadays as you have reported, the o0ld warhorse is being taken down to the knacker's yard for final disposal and boiling down for glue. What have they left from the old HP portfolio once the 3000 is got rid of? The answer is the hardware. Now we all know that the margin on hardware in the Wintel (sorry about that, I had to go and be ill) are minute and are changing all of the time. Of course the Dell model of production is the ideal with very very little (or no) parts inventory at the end of the year, but there are few manufacturers who can do this. I strongly believe that Ms. Fiorina has done the company no service at all by buying out/ merging with Compaq.
Reading "The HP Way" by Dave Packard is like (to quote Mr Connelly) "Running through a swamp with your wellies full of vomit" but it does get across the original ethic of HP. They were hardware men through and through. They developed the inkjet technologies of today, the humble LED, the basis of the laser printer and so on. Now this has been got rid of and they will soon find that the cupboard is bare of good money-making ideas. They have still got the HP/UX boxes and software but not many people have much faith in the longevity of this operating system. The same goes for Open/VMS (which had a makeover not so many months ago with features that have been in the MPE world for YEARS). Also, what about the TRUE64 world? Is that due to go the same way?
[name and address supplied]
Don't go deifying Ken Olson. Digital was a great company from its start until about 1972, but then it fell into irreversible decline because of his consensus management policy. This led to innumerable small internal fiefdoms asserting their power, making it nearly impossible to get anything done because everybody had to agree on everything. All you had to do to have power was to withhold your agreement. Sure, a few things got pushed out the door, but the high margins just hid bloat, they didn't support R&D as much as they could have done. The main product was a blizzard of internal memos - so many that most days could be spent just responding to them. Parkinson's Law run amok.
Everybody around here knew about Digital. A tiny, continually shrinking, and finally nonexistent handful of top guns and a huge swarm of drones. By the 90's you didn't dare put Digital on your resume if you wanted to get a job in the Boston area. Recruiters didn't want to talk to you. The Sun and HP facilities in this area went the same route, though not as egregiously as Digital.
By the mid-to-late-80's nearly all the top players at Digital were gone and the corporate culture was a morass. People undoubtedly took ownership of projects, but usually not more than once, and the universal buy-ins (meaning everybody remotely connected to a project had to be fully informed and had to agree in detail to every aspect) slowed progress to a crawl. There's also the question of whether the projects were worth doing, they were so off-the-wall.
The only penalty for failure - and failure was concealed in a variety of ways including a project being subsumed into something else, or reorgs - was that you'd remain where you were, just bitter. Or if you were enough of a pain-in-the-ass, you'd be kicked upstairs. Parkinson was a remarkable man. or really familiar with bureauracracy, and the Peter Principle was confirmed. So many people were at their level of incompetence.
At its peak, I think Digital had 150,000 plus employees when they really needed only around 30,000. Then big job cuts started under Palmer, Ken Olson didn't have the heart to do it, but they were too small and too slow to save the company. Not only that, but the cuts just reinforced the cronyism that was rampant by then. If an incompetent manager had to cut people, he wouldn't cut the deadwood because those were most likely his/her buddies.
I saw much the same thing at Prime Computer, Digital's biggest mini-computer rival. That I also had a front row seat for. Cronyism was more of a factor there, and poor decisions based on linear thinking - we'll never abandon PRIMOS for UNIX, but that's another story.
When Intel took over the Digital fab in Hudson, MA, I was with the Alpha design team there at the time and I was impressed with the way Intel came in and tried to turn around and ailing plant. Don't know if they ultimately succeeded.
Microsoft wouldn't go under because of a poor product, but poor sales would hurt their stock just like EDS or Apple or any other company. And they'd probably keep plugging away at it until they got it right. MS is a tight, well managed company, at least here.
A very good article and largely spot on. Although I differ on a few subtleties. I'm head of technology for an investment management firm, and I see the "flow of funds" all the time.
You are absolutely correct to say that capital goes careening around like a pinball in the pursuit of the largest return in the shortest time. But, that is of course what lenders of capital want. Who cares about the underlying validity of the process, much less the moral issues. Increase your wealth and get out.
Within some guidelines of "reasonableness" this doesn't offend me. It's logical for people to want to make as much money as possible, assuming they don't break the law, (which law?), in the process. We all know that the line for business turpitude is very smudged from being trod on so frequently. But that doesn't negate the fact that there are people with
innovative ideas, both for themselves and society as a whole, and there are people with money. Of course they're usually not the same people. So there has to be some sort of "market" to bring these types together.
The fact that the "money people" frequently get it wrong is to me what's good about capitalism. If you're over hasty and stupid you'll be punished financially. Hopefully you don't take a lot of "innocents" with you. But those who wish to "play" in capital markets, i.e. make money without having to do much work, should be aware of the risks. Greed sometimes gets rewarded in big numbers, but also equally gets slammed for excess.
I don't believe that the current climate of money running from technology will last too long once capitalists find a "model" in which they can make money. The problem is that historically much money has been made from constraining or limiting information. In the current state of technology that is becoming harder to do, hence that model is dying. Even Microsoft, who has milked that "cow" as well as anyone is starting to figure that out
and is adjusting "policy" accordingly.
Money flows usually represent the excesses of society, not its core values. Water sloshing about in the tub is fine and long as you have no problems with getting wet.
Dimensional Fund Advisors
Much to agree with there - but does greed get punished "equally"? It took the robber barons era for the government to create the antitrust legislation, and the 1929 crash to create SEC.
One progressive favorite is a tax on currency speculation. Dave Loveluck in Snohomish, WA suggests a five-year minimum hold period for stocks. But how would could that not penalize investors who were left in holdings in a company that might be shown to be criminally negligent or simply incompetent, long before the five years was up? Then there's the suggestion that investors shouldn't be able to profit for a certain length of time.
But don't blame the VCs, writes one reader:-
A very good and insightful article. To add to your thoughts, the larger technology companies certainly have themselves to blame as much or more than the venture capital system. Some companies (you named a couple) made a commitment to pour earnings into R&D with the hope that they would generate far greater earnings in the future. Conversely, those companies that dished out lavish sums to their execs and/or used their earnings for stock buy-backs are the ones that later ended up at the mercy of the VC's.
When companies gain the experience and knowledge to guide themselves to do the right thing, they worry less about Wall Street and more about their own projects.
One question: if we have an oversupply of broadband, why are prices rising at twice the rate of inflation and why is there still a monopoly in the deployments (ie most can't get dsl but can get cable only)?
I suspect because there isn't real competition between cable and DSL providers. It's often just one per area. Here's some more on "the only value you can put on a banana is a bruise":-
My thoughts on the commoditization factor:
A network externality means that users *want* commodities. The externality means that users profit by having bought the most popular thing. Hence, even without the financial markets, commoditization is the natural endpoint for some parts of the tech business. The promise of the web is to allow standards to provide the positive externality of compatiability without the attendant commoditization. Hence Redmond's interest in using .NET to stop the J2EE/XML trend.
But in the meantime, commoditization works in tech because it is what we, the users, want. In other areas, low margin doesn't always garner financial market rewards. Whole Foods, for example, commands a P/E ratio of 38, by adding more value to their fancy bananas, than a bruise.
Safeway, who has the adding bruises bit down to an art, has a P/E of 8. But if you had to change out your kitchen in order to eat a Whole Foods banana, then Safeway would be would be where we would all want to shop.
Chris van Loben
The analysis you give of the effect of speculative capital in large scale computing businesses is very good, but it omits a crucial element: the end purchaser. However great the value supply chain, if there is no end value, there ain't no chain. And that's what we're calling a recession in the computing industry, the collapse in end use demand.
One of the major analysts produced an industry survey that suggested that in 2001, one third of all first-time PC consumers, purchased their PC for one reason: Napster. Now, however righteous the case for strangling Napster, strangling it probably cost one third of the first-time consumer marketplace. Without those punters picking up the iron at the malls and warehouses, the whole economy of scale thing, for all PC producers, has gone very hostile, with pay-back into the margins on the commercial side.
It's like the two bete noirs of Microsoft: activation and predatory licensing. Activation on the macro scale is like crack addiction in the cash flow timeline, you have to keep having more, harder and stronger. At first it kills off the underbelly of your market, taking out the soft option purchasers who migrate from stolen to legitimate. So in order to enforce better volumes, you ensure even tougher activation. Slowly the alienation sets in among power user communities, which is addressed by even more stringent hostility towards those purchasing power users. It's a self reinforcing downward spiral, driven ultimately by shareholder earnings, a kind of inverted Ponzi scheme killing off tomorrow's revenues to fix the hole in today's.
Licensing is even more toxic over the long haul. It spikes early revenues in the first two years but after that the 60% who didn't buy in, sit on their existing systems for far longer than they otherwise would. This has the effect of slewing the Windows deployment base, as older installs persist longer than they otherwise would: delaying that crucial upgrade in the developer skill base, as old skills remain viable far longer; making the burn factor of switching to Linux increasingly less onerous over time, as the burn factor of upgrading within Windows grows worse the father back the legacy deployment becomes.
What you're seeing happening in today's major vendors and producers is the after-blast following overly optimistic financing by precisely the same entities that blovated these large concerns in the first place at the cost of supplier diversity. And thereby driving what is felt to be 'a reality' from the investors' perspective, into the industry. Thing is, that 'reality' IS killing off the goose, the gander, the homestead and the ranch, and it's A Good Thing, dinosaurs are always at their biggest just before the extinction. Whoever would have believed in the decline of IBM before it happened?
The question is: do we really need Intel, Microsoft, Dell, HP-Compaq and all the other hegemons that have so badly distorted the computing market? Hell no. What is currently passing itself off as R&D is not any kind of attempt to bring real value or genuine innovation to the end consumer, simply an attempt to create some of kind of novel market that can sustain the scale and the cash cows created during the bubble. Innovation requires an unconstrained marketplace in which newness can compete with oldness; in a market with no supplier diversity, innovation cannot compete with legacy - this again was proven by IBM's incapability of allowing the early PC to compete with its mainframe market.
Let the power fall. The only way for the mass majors is down. It will not be an overnight catastrophe, it will be preceded by a further curtailing in supplier diversity, it will come as an inexorable anaemia of revenues as the 2003 benchmark for a step change in the server (and now desktop) approaches, and after that the 'inverse Ponzi' effect will split these companies open as the expected re-emergence of their bull markets never happens. It doesn't matter how huge the economy of scale if the demand isn't there. And scale in R&D only determines speed, not outcome.
Sooner or later capital will speculate on a better gamble in a better marketplace with a better industrial model.
Real innovation always started small, and it always will. Diversity and innovation will always return to hegemonist marketplaces through demand induced crisis - this is basic Keynes for crying out loud, only there ain't no government to synthesise economic growth.
S G H Houbart
I congratulate you on your timely, incisive analysis of HP's ongoing business plan and its uncanny resemblance to the Fall and Decline of the DEC-an Empire. I also welcome your attempt at wider analysis. While risky, the courage to examine new ideas always makes society's discourse richer. Thus, I am a little tremulous to point out that your analysis contains two misunderstandings that undermine your conclusions.
1. Economic advancement is the result of specialization of labor, not investment capital.
2. People innovate, not companies.
Allow me to address each in turn.
1. The key to economic advancement is specialization of labor, not investment capital.
The key Adam Smith's "The Wealth of Nations" is that as individuals and groups of individuals "specialize." That is, they become particularly good at what they do, such as making nails in Adam's example. If each person concentrates on what they are best at, society advances. Now let's look again at your analysis. You state "Fortunately a few companies defy the logic: Sun, EMC and Apple amongst them" and then you describe how each is a specialist. Which doesn't defy logic at all -- it's exactly what Adam says should happen. The deeper you look, the more pervasive Adam's logic becomes. Dell knows it has no R&D -- that's why it's been cozying up to IBM. IBM's "barbell" strategy is to specialize in R&D and Systems Integration, and leave manufacturing to "distributors."
This leaves Dell and whomever has licensed IBM disk drive technology to specialize in the quite different skill set needed to crank out cheap units quickly. Comfortingly, applying Adam's logic suggests that you are right: HP's current strategy is nuts. As you pointed out, they have no compelling specialty.
2. People innovate, not companies.
You lament that HP and DEC, among others, are no longer the innovators they once were. By focusing on corporate organization you have missed the picture entirely. These current companies are merely the legal continuations of the innovator's workplaces. The innovative _people_ left long ago. How much innovation has Palm done since Trip Hawkins and Donna Dubinksy left ?
I appreciate your attempt to analyze the evolving ecology of the tech world. Such a task is never easy, and can be approached from many vantages.
On the other hand
Great article. Todays capitalism is not that envisioned by Adam Smith.
The cancerous form alive today runs far deeper than the technology industry.
We are in big trouble.
But frankly, I don't think that the whole problem can be laid at the feet of the financial goons. It seems to me fair to point out that Intel, Microsoft, AOL, and Amazon drove very hard for market dominance in preference to thoughtful design based on careful research. I don't think those guys were forced into it by greedy VC's -- rather it was their basic nature, and they hooked the capitalists by their greed.
So the really interesting question, to me at least, is whether there is a way out of the trap: whether good design, careful implementation, and meaningful customer service can be successful in a modern economy. And it may be that the Japanese auto makers are a workable model -- though they do depend on a financial time horizon beyond the next quarter.
Finally, what did Olsen himself have to say about Digital after he was ousted in 1992. Not too much, as when interviewed, he's declined to comment on his successors. He read about the Compaq acquisition in the papers, he said.
From the Boston Globe [link]
Short-sighted investors will do anything to have $1 more at the end of the quarter, the ousted founder of Digital, Ken Olsen, told a forum of about 300 entrepreneurs in March 1994.
"The worst thing one can do is read the front page of the business section of The New York Times to figure out what they are. I guarantee every one of those is a failure," he declared.
"I say never go public with your company. The only trouble is there's often no choice," Olsen told the entrepreneurs. Going public, he says, means putting the company's fate in the hands of often short-sighted investors. "Owners today have no loyalty to the company. They don't care about the company or the country. They would do anything to have $1 more at the end of the quarter," he said.
As for his tenure at Digital, Olsen said "I was beat up unmercifully" in the press and by analysts "for not having mass layoffs."
"I've thought about that many times, but I never say anything. My history has always been to solve problems. We rarely had the fastest computer. We did take the best care of customers.".
That's no gwana-gwana. ®