Join 3,430 readers in helping fund MetaFilter (Hide)

Tags:

A financial engineering operation masquerading as a technology company
February 9, 2014 8:51 AM   Subscribe

A series of columns by Robert X. Cringely about the decline and impending demise of IBM

posted by JeffL (104 comments total) 95 users marked this as a favorite

 
IBM was never a technology company, it was a marketing company that used tech as leverage. At some periods there was so much extra money that indulging in "research" was an extra benefit.
posted by sammyo at 9:05 AM on February 9


One thing I like about metafilter is that 3 people have marked this as a favourite (i.e. it is something they want to read and perhaps return to), but nobody is posting until they have read it. (Well except me obviously.) metafilter is a rare and wonderful thing on the knee jerk internet.
posted by EnterTheStory at 9:06 AM on February 9 [10 favorites]


I worked for Large Blue from '99 when they bought the company that I worked for until 2005 when they finally noticed that I was still working there and laid me off. In those six year, I never did figure out what IBM did for a living. They certainly weren't even slightly interested in developing software.
posted by octothorpe at 9:17 AM on February 9 [14 favorites]


I often wonder if gigantic corporations like this, when they're nearing the end of their existence, just make a game out of how long and how much money they can squeeze out of smaller and smaller customer bases until it's all gone? These articles certainly make it look like IBM has become a dying carcass that executive vultures keep picking at until everything's gone.
posted by xingcat at 9:19 AM on February 9 [1 favorite]


I had been following this, so the latest Cringley column about IBM selling their Intel business was the one I hadn't seen yet. A nice complimentary article to the "IBM to customers: Your hand is staining my window" link is a Jeff Matthews column from January - IBM: "I’ve Been Manipulated"?. My employer is a very large IBM customer and these Cringley articles have been read and discussed by everyone who has to deal with IBM on a regular basis. All of the stuff about services rings true. I don't agree with the last article about IBM being doomed because "they need to learn how to operate in a commodity business" rather than selling it (the x86 business) off. They are essentially turning into a hedge fund and their product is Earning Per Share. Given their EPS roadmap that Matthews talks about, I don't know that selling the x86 business was the wrong move. Being a technology company that produces technology products isn't necessarily part of that roadmap.
posted by BlueTongueLizard at 9:19 AM on February 9 [3 favorites]


IBM was never a technology company, it was a marketing company that used tech as leverage.

Herman Hollerith and his staff, playing with pools of mercury to get punch cards to register, were in marketing?
posted by wobdev at 9:20 AM on February 9 [24 favorites]


My uncle worked at IBM up until the 90's. I'm not kidding that he was probably one of the smartest people I had ever met, and the man bled blue. He was very proud to be working in Yorktown. I'm not sure people appreciate how far IBM has fallen at least in terms of work culture.

IBM seems to believe it is cheaper to replace a skilled worker with two or three unskilled workers to do the same job. That is like hiring nine women to make a baby in one month. Today at IBM the US workers who try to save the business are the first in line to lose their jobs. Management accountability is gone. The people who mess up get to keep their jobs; and those trying to retain the business lose their jobs.

Devastating. I guess I'm glad he's didn't live long enough to see this.
posted by phaedon at 9:22 AM on February 9 [11 favorites]


Cringely? Really? If he's saying IBM is dying, I'm off to buy IBM stock, because that means they're about to make a huge comeback.
posted by entropicamericana at 9:24 AM on February 9 [15 favorites]


This is the inevitable result of the growth of the executive class, and its loving embrace of speculation. These are by and large people who suffer from two fatal personality defects: they have never worked for a living; they believe that things they don't understand are not important. The combination results in merciless cuts to the livelihoods of working people (for whom they have no empathy) and shipping work abroad to people and places not competent to do it (because anyone can do "work," whatever that is). I don't know where these executives think they or their children will live after they've turned our society into a smoking pit mine. Switzerland, probably.
posted by 1adam12 at 9:36 AM on February 9 [29 favorites]


At one point IBM workers sang paens to Big Blue.

Some good it did them.
posted by wobdev at 9:38 AM on February 9 [3 favorites]


I hear their analytics division is doing quite well. Also, Watson!
posted by zscore at 9:40 AM on February 9


BM seems to believe it is cheaper to replace a skilled worker with two or three unskilled workers to do the same job.

One of my last tasks at IBM was to train the team of ten or so people who were going to take over the project that two of us were doing part-time.
posted by octothorpe at 9:43 AM on February 9 [19 favorites]


my brother still works there in rochester - started in the late 70s - and for the last few years, he's been waiting for the axe to fall

he can get early retirement and has lived well within his means, so he's not real concerned about it

he doesn't really say much about it because he can't - but things are pretty messed up
posted by pyramid termite at 9:47 AM on February 9


Large corporations need to adopt the "George Costanza Rule" and just know when to go out on top.

Instead of making a big deal about selling off parts of your business, you're better suited to follow the Apple model and stop highlighting a particular product line and just let it coast along.

Of course, this implies that you, as a corporation, have already jumped onto 'the next big thing' and are currently riding that wave. If not, then the last person in the building, please turn out the lights...
posted by tgrundke at 9:48 AM on February 9


I'll always remember the nice IBM rep who stopped by every time I shuffled across the carpet and fryed my Lenovo, er Thinkpad network interface with a static discharge.

I'll also remember doing WebSphere contracting and taking two freaking weeks to figure out how to import a static web site. (WebSphere is/was an extremely convoluted XML-centric version of JBoss that cost around $100k per CPU per year back in the day, and required an army of IBM contractors to develop for and maintain it because it was evil and opaque -- a perfect product for federal IT projects in other words.)
posted by RobotVoodooPower at 9:52 AM on February 9 [1 favorite]


It looks to me like IBM is gearing up to build a gigantic Watson business. If it's nearly as good as it looks, in 15 years every business will have one and in 20 every home will too.

Selling commodity Intel servers is what's keeping them afloat? I can't see why that shouldn't be done in India or China, what exactly are Dell and HP bringing to that party?
posted by Nelson69 at 9:55 AM on February 9 [2 favorites]


They're going into Africa in a big way right now, after the Nairobi Innovation Center, they've just announced new ones in Lagos, Casablanca and Johannesburg, but as the BBC's own Akwasi Sarpong tweets, there is a long way to go before they can improve their image among thinking Africans.
posted by infini at 9:56 AM on February 9 [3 favorites]


I've been thinking over the past several years, as IBM sold off more and more to Lenovo, that they seemed to be getting away from their core enterprise. Lenovo bought IBM's pc business in 2005, then just recently bought IBM's server business as the last link states and interprets. Now, Reuters is reporting that IBM is considering selling its semiconductor business, and has hired Goldman Sachs to find a buyer.

It seems like a shame, because there's so much potential for further advances in computing technology. It won't be IBM that brings the first quantum computer to the mass market, it will be Google. There are, after all, lifecycles to businesses.
posted by clockzero at 9:57 AM on February 9


This seems like the behavior of any other behemoth corporation burdened with the onerous task of forever increasing efficiency by lowering costs through whatever means are at its disposal – in this case, reducing labor costs by the tried-and-true method of getting rid of highly paid senior employees and replacing them with low-cost junior employees or outsourced labor. That's a sleazy move, but by no means unusual in the corporate world, nor indicative of the sort of executive incompetence Cringley is alleging. It has always been my subjective experience that outsourced contract workers are inferior to seasoned domestic workers (at least, in the software world), but is that assertion actually quantified anywhere? Or are we just supposed to assume Cringley is an expert on this topic and doesn't need to cite any sources or research or anything? If IBM is knowingly reducing the quality of its services by outsourcing them to lower-cost and lower-quality workers and operations abroad, then that's behavior deserving of the author's admonishing tone, but it just sort of seems like he's making shit up, pinning the blame on these supposedly unqualified people in BRIC countries who are destroying the company.
posted by deathpanels at 10:01 AM on February 9


Just for the record, about IBM's past, none too clean a punch card there either.
posted by infini at 10:02 AM on February 9 [1 favorite]


Fulfilling customer requirements is a weapon at IBM

This is how you make billions of dollars as a contractor for the federal government. It's precisely why healthcare.gov was the massive clusterf*ck it turned out to be: there were very, very few contractors capable of turning in a conforming bid, and those that did knew exactly how to game the system such that they could keep all the money they were paid regardless of whether the product actually worked.

It's not just IT either; a huge amount of this happens in defense contracting. The Coast Guard once had the bright idea to commission a project where you took a cutter, chopped the end off, added a couple dozen feet of length, and called it a bigger cutter. If this sounds like an absolutely horrible idea, that's because it was. The first time they took one of these puppies out for a cruise, it literally broke in half and sank. The Coast Guard sued, saying the shipyard made false statements about the project.

Turns out what happened--I have a relative in the shipbuilding business who knows about this first hand--is that the initial specs had an error which the shipyard reported to the Coast Guard. The error made the whole project non-viable. But the Coast Guard didn't have the institutional flexibility to either can the project or at least take it in a different direction, so they just told the shipyard to make whatever fixes it could and damn the torpedoes. With predictable results.

The federal judge who dismissed the case said that the Coast Guard (by way of the DOJ, who were the ones who actually brought the case) didn't even allege that the shipyard had done anything wrong. They did exactly what they were told to do. The fact that it was an egregiously stupid idea was the Coast Guard's fault, not the shipyard's.

Of course, that kind of bullshit doesn't work in the private sector. If a client doesn't get a working product, they go somewhere else. But IBM, like CGI Federal of healthcare.gov fame, is for a variety of arcane reasons, one of the few companies that bothers to submit conforming bids for major projects like the Pennsylvania benefits disaster. Something almost identical happened in Indiana. The states are never going to get that money back, because the contractor did everything the contract required of them. And because of the laws that govern government contracting, agencies can't simply ignore conforming bids based on bad experiences in the past. If IBM (or some other company with a bad track record) submits a conforming bid, and it's the lowest bid, it's very, very difficult for agencies to pick a different one.

Which is why companies like IBM continue to exist, i.e., there is a massive source of revenue for companies large enough to bid on these projects and there's no penalty for screwing them up.
posted by valkyryn at 10:02 AM on February 9 [45 favorites]


Selling commodity Intel servers is what's keeping them afloat? I can't see why that shouldn't be done in India or China, what exactly are Dell and HP bringing to that party?

What's been keeping them afloat, in large part, is their services division. That's the one where they have moved to a model of charging big companies $200 per hour while paying someone in a low-wage country $10 per hour to do the work. There have been several high profile failed projects with that model, and at some point companies are going to stop paying IBM all that money for stuff that doesn't work.

Companies that want to go the cheaper offshore route can just hire Wipro or similar companies themselves, and cut out the expensive middleman.
posted by JeffL at 10:03 AM on February 9 [1 favorite]


They're going into Africa in a big way right now, after the Nairobi Innovation Center, they've just announced new ones in Lagos, Casablanca and Johannesburg, but as the BBC's own Akwasi Sarpong tweets, there is a long way to go before they can improve their image among thinking Africans.

Now, this is a very interesting story. Personally, given my background in international development and in social science, I'm very skeptical about the breadth of applications that the Watson technology can really help with in a meaningful sense, but introducing powerful information technologies will probably spur further communication infrastructure investment connecting Africa with the rest of the world, which probably will help potentiate further investment and development, and that could work out in a beneficial way for African people. It makes sense, on a certain level: IBM is pivoting (or trying to pivot) to a totally new market and new products/services here.
posted by clockzero at 10:05 AM on February 9


Of course, that kind of bullshit doesn't work in the private sector. If a client doesn't get a working product, they go somewhere else.

Exactly. But is there enough government work to keep IBM's services division afloat? And they're not guaranteed to get all of that anymore - as in the recent case where Amazon beat out IBM on the CIA contract
posted by JeffL at 10:10 AM on February 9


Cringely? Really? If he's saying IBM is dying, I'm off to buy IBM stock, because that means they're about to make a huge comeback.

Maybe you're confusing him with John Dvorak, who has been literally as wrong about Apple as it is possible to be, for decades?

And I do have to admire this line from the first post:

IBM seems to believe it is cheaper to replace a skilled worker with two or three unskilled workers to do the same job. That is like hiring nine women to make a baby in one month.
posted by Halloween Jack at 10:23 AM on February 9


And I do have to admire this line from the first post...

Halloween Jack: This is an oblique reference to Brooks's Law.
posted by ZenMasterThis at 10:32 AM on February 9 [2 favorites]


It makes sense, on a certain level: IBM is pivoting (or trying to pivot) to a totally new market and new products/services here.

clockzero, what you postulate makes a lot of sense when seen in isolation but in the context of technological introduction, adoption and innovation across Africa, IBM is really one of the last to enter the fray. They're probably going to rely on their Airtel contracts from India to get introduced to services contracts, and more likely to go after government big data stuff than any pivot or change wrt to mass majority social and economic development.

links on request.
posted by infini at 10:41 AM on February 9 [1 favorite]


I also tend to use favorites as a "read later" mark.
posted by destro at 10:44 AM on February 9 [2 favorites]


Also to consider: Watson may, at this point, just be smart enough to predict the stock market's behavior well enough to keep their cash on hand doubling every few months...or days.
posted by sexyrobot at 10:44 AM on February 9


There are two issues at hand
1)IBM's insane accounting and the culture surrounding Wall Street and "Hitting the Numbers" - which is obviously terrible and indefensible
2) IBM's transition (probably under Gerstner) from a Technology company that viewed R&D as the best investment for them to make to sustain growth and profitability to an extremely well run industrial conglomerate.

These days IBM is fundamentally the same thing as 3M, Emerson, GE, Honeywell. And that's fine. But really its a big pile of businesses (Some of which require reinvestment in the form of R&D spending, so of which do not) that generate cash that the parent company reinvests by buying new businesses. The parent company also sells businesses they think aren't as good any more.

And while they occasionally have a dud here or there - they are actually pretty good at it. I'm opinionless on the price other people are willing to pay for the shares - but it is what it is.
posted by JPD at 10:48 AM on February 9 [7 favorites]


Also to consider: Watson may, at this point, just be smart enough to predict the stock market's behavior well enough to keep their cash on hand doubling every few months...or days.

Thats.....probably not what's going on here.
posted by JPD at 10:50 AM on February 9 [15 favorites]


In 1980, my Accounting job was converted from a manual "pegboard" system to the small company's new IBM System 360 "Mini" Computer (at the time "mini" meant it was smaller than TWO refrigerators) using 8-inch floppy disks. In 1986, I joined a company that had begun supplementing the desktop terminals hardwired to the company's mainframe with a few IBM PCs and got access to one of them. At my next new job in 1992, I was shown to a desk with a Gateway 2000 PC. Twenty-two years ago. IBM still exists?
posted by oneswellfoop at 10:55 AM on February 9 [3 favorites]


The reason IBM can’t deliver is also explained well by Steve Jobs. It’s IBM’s maniacal fixation on process, once a strength but now a cancer.

“Companies get confused,” Jobs told me. “When they start getting bigger they want to replicate their initial success. And a lot of them think well somehow there is some magic in the process of how that success was created so they start to try to institutionalize process across the company. And before very long people get very confused that the process is the content. And that’s ultimately the downfall of IBM. IBM has the best process people in the world. They just forgot about the content.”


I just want to point out how brilliant this insight is. You know what you can turn into a process though? Growth through M&A. And that's what IBM does.
posted by JPD at 11:03 AM on February 9 [18 favorites]


These days IBM is fundamentally the same thing as 3M, Emerson,...

Well, at least in the case of my former employer Emerson, maybe not quite the same thing: IBM 39, Emerson Electric 2
posted by JeffL at 11:09 AM on February 9 [2 favorites]


X86 server industry is doomed, everyone knows it. Data centers are moving to ARM and custom systems that look and operate noting like an old fashioned PC while hosting zillions of virtual machines.

IBM can try to break in on the ground floor of a low margin field already crowded wirth competitors, including the largest potential customers like Facebook, Google, Rackspace and Amazon who are designing their own hardware... Or it can sell bespoke silicon designed for heavy lifting tasks that don't mesh well with commodity systems. (Map-Reduce can't solve every big data problem).

IBM has never taken a loss from its mainframe division, just saying.

Identifying a changing server marketplace before it kicks your ass and eats your profit margin is one of the few smart things they've done recently.
posted by Slap*Happy at 11:26 AM on February 9 [10 favorites]


Good news, everyone! My home town is fucked.
posted by maryr at 11:36 AM on February 9


I'm puzzled by the idea that clusterfuck projects only happen when The Government is the customer. In fact, I'd suggest the vast majority of awful work is done on behalf of large corporations, who take delivery of awful products and systems all the time. I'm working on one now, where our sweetly working, perfectly capable system is being migrated to a completely unsuitable environment because someone bought the CTO lunch.
posted by maxwelton at 11:47 AM on February 9 [19 favorites]


(Honestly, I just hope IBM holds out until my dad, who was lucky enough to be old, gets his all his pension money out. ...So, I suppose that means I want them to hold out a good long while as I hope my father continues to draw pension for years and years to come. Go Big Blue?)
posted by maryr at 11:48 AM on February 9 [1 favorite]


I emailed this thread to a friend whose wife is actually relatively senior in one area, and his response started "Ha ha, you won't find any group of people more pessimistic about IBM's long term viability than IBM's own employees." He went on to say that while they certainly were trying to outsource to India as fast as possible, it was such a clusterfuck that they fired an executive in charge of it.

Cringely makes it sound like they're screwing their customers quite expertly, but plainly, the insides match the outsides for crappy, overpriced service delivered to the letter of the plan, in total violation of the spirit.

Is there an accepted term for companies like IBM now, where their core competency is a shell game played with spreadsheets? "Cell manipulation vehicles"?
posted by fatbird at 11:54 AM on February 9 [1 favorite]


Maybe you're confusing him with John Dvorak, who has been literally as wrong about Apple as it is possible to be, for decades?

Dvorak is a troll; I have the impression Cringely actually believes the nonsense he writes.
posted by entropicamericana at 11:58 AM on February 9 [1 favorite]


Dvorak is a troll; I have the impression Cringely actually believes the nonsense he writes.

What's nonsensical about what he's written about IBM? Even the one IBM-related thing of his that I don't really agree with (the most recent column about IBM selling their low-end server business) doesn't strike me as "nonsensical."
posted by JeffL at 12:13 PM on February 9


I was "downsized" i.e. "invited to leave the business" or "fired" from IBM in 1999. It's astonishing to look back at the damage that IBM did to employee morale in the 1990's and not see it as intentional now.

There were absolutely incomprehensible personnel moves, wholesale shuffling of employees off to paper subsidiaries to break and then raid pension plans, and employee rankings where managers were forced to grade 20 percent of their employees as "undesirable", despite years of hiring the top 1 or 2 percent of available personnel.

Fuck them. Fuck them with a 4x4 wrapped in barbed wire covered in cayenne pepper powder. I hope the I.B.M. sinks faster than the Titanic, and with no lifeboats for upper management.
posted by pjern at 12:34 PM on February 9 [11 favorites]


What's nonsensical about what he's written about IBM? Even the one IBM-related thing of his that I don't really agree with (the most recent column about IBM selling their low-end server business) doesn't strike me as "nonsensical."

Never mind, Cringely is a prophet and a seer, okay? I'm not going to write a dissertation to defend a goddamned quip.
posted by entropicamericana at 12:36 PM on February 9


IBM was never a technology company, it was a marketing company that used tech as leverage. At some periods there was so much extra money that indulging in "research" was an extra benefit.
This is about as wrong of a comment as I've ever seen on the Blue. Are you confusing IBM with pets.com or something?

Anyway, my entire company's backend is based on AS400/RPG (no not the fun kind of RPG) and I hate the tech so much even though I know it's more economical to stay with it. And there's tons of companies out there still reliant on RPG and Cobol. From my understanding IBM will still be making mid and mainframes so we'll still have our hardware supplier. Lenovo ThinkPads have been OK but I hate what they've done to the classic keyboard.
posted by kmz at 12:41 PM on February 9 [3 favorites]


The only clear thing to emerge from this is that the idea of stock trading (ie, holding shares in a company for short-term profit) is completely contrary to the good of the businesses being traded (and more importantly, the economy in general and society as a whole)--unless, of course, you are directly involved in profiting from the trading (which happens to include the highly-paid executives of the businesses involved).

The only thing "the market" rewards is (insane) growth, and the truth of it is there are more opportunities for good businesses which exist as essentially static entities, delivering a reliable product to a known customer base, than there are for wildly growing companies (many of which are only growing inasmuch as they're burning through borrowed money with no plan for how to pay that back). The idea you can grow indefinitely is bizarre. Anyone who believes it is an idiot.

As we take more and more money from the largest potential base of customers, why aren't dreams of infinite growth met with hysterical laughter? It's one thing to dream of everyone in the world owning one of your widgets, that's a big (but finite!) market. It's another to realize that your potential customer base is the 1000 people in the world with actual money, most of whom already have something which does what you claim to do.

Anyone who suggests "they should run that like a business" (typically "government") has never actually been inside of a business with stock traded on the market. Government already is more efficient than most large businesses at getting stuff done.
posted by maxwelton at 12:44 PM on February 9 [16 favorites]


I haven't heard the words AS400 since 1991 when my college boyfriend trained on the midrange and departed for the yewnitedstatesofumrika from chennai.

*cue violins and nostalgia*

i rather enjoy the current lenovo keyboard on the edge, i'm typing twice as fast now
posted by infini at 12:47 PM on February 9 [1 favorite]


I don't agree with Cringley's "IBM is doomed" article because IBM is spinning off its x86 server business. Just like it wasn't doomed when it sold off its Thinkpad and desktop business years ago. IBM has been in the services business for years and if one of their clients need x86 servers then they'll just resell them a Dell/HP/Lenovo/brandX servers.

If anything this is like them selling their buggy whip business as the see that there's cars (ARM-based servers) coming along. Dell and HP have bought services companies to try and drink from IBM and other services companies' milkshake.

In the long term the only one that will be making money on x86 serves are Intel and Microsoft. The reason Google and FB make their own servers from commodity components is because the OEM logo on the box isn't worth the extra cost.
posted by birdherder at 12:51 PM on February 9 [3 favorites]


I really don't understand selling the server business. IBM sells complete systems, including hardware, custom software, and support; there's a lot more value in saying "if your database ever breaks for any reason we'll fix it" rather than "well, it's a hardware problem so you'll have to contact your hardware company." And no one can afford to run everything on IBM's remaining exotic server lines.
posted by miyabo at 12:53 PM on February 9


IBM has been in the services business for years

He's not saying they're doomed because of the sale, he's observed that after transitioning to a service model under Gerstner, they've systematically removed or destroyed every part of the business that is capable of producing value, while leaching the value out of the relationships they've built up, all in the name of keeping the share price up and increasing EPS. In other words, selling the server business is of a piece with their auto-cannibalization, rather than the remaining nub of the original business that could have been the foundation for a turnaround.
posted by fatbird at 1:02 PM on February 9 [5 favorites]


If the late CK Prahalad were here, he'd be mumbling about 'core competencies' being essential.
posted by infini at 1:04 PM on February 9


IBM seems to believe it is cheaper to replace a skilled worker with two or three unskilled workers to do the same job.
One of my last tasks at IBM was to train the team of ten or so people who were going to take over the project that two of us were doing part-time.

Well, the Mythical Man Month was literally written about a software project at IBM. I guess they never learned.

It has always been my subjective experience that outsourced contract workers are inferior to seasoned domestic workers (at least, in the software world), but is that assertion actually quantified anywhere?

I can't quantify it, but I do have some social arguments to justify it the observation:
  1. When companies offshore, it's almost always done for cost-cutting reasons, rarely for quality*. There's also brain-drain from lower-paying countries to better-paying countries, because (I assume) more talented people have better access to the opportunity to emigrate. Since the companies are approaching offshoring as cheapskates, they're only going to get the people who can only work at the low end of the pay scale for reasons of inexperience or incompetence.
  2. A relatively higher proportion of support and maintenance work means offshore workers have less opportunity to develop their skills.
  3. Huge time-zone differences that preclude effective communication and cooperation.
* e.g. off-hours support
posted by cosmic.osmo at 1:39 PM on February 9 [4 favorites]


Slap*Happy: "X86 server industry is doomed, everyone knows it. Data centers are moving to ARM and custom systems that look and operate noting like an old fashioned PC while hosting zillions of virtual machines.

...

Identifying a changing server marketplace before it kicks your ass and eats your profit margin is one of the few smart things they've done recently.
"

Over 15 years ago, Clayton Christensen published a framework (a process, if you will) by which newer but inferior technology slowly invades higher margin markets. Chips designed for cash registers and calculators slowly find applications in desktops, and then in servers, as they become more reliable and add new features. Intel did it, and now ARM is doing it. As you say, this is obvious to everyone. I wouldn't say the x86 server industry is doomed, but they will need to be comfortable on much thinner margins.

The one minor challenge one is how to place IBM's Power chips in that spectrum. In clear defiance of Clay's observation, they tried to push Power into desktops, and that lasted surprisingly long before being unseated by Intel in their one flagship desktop customer. I don't really understand it, but somehow Power is good for embedded applications and super high performance. As it happens, the intersection of embedded and super high performance is gaming consoles, and Power is absolutely killing it there, being featured in all three major consoles. I can't imagine P-series as being very popular for anything other than continuous integration farms for game developers.

I could believe that IBM having multiple product lines makes it hard to marshall resources internally well, having a Linux Technology Center with a separate budget from the AIX division, separate x86 server and the Power server divisions, etc must make for some fun times. But with rumors IBM looking to sell its semiconductor unit, I think the game plan is a bit different than 'lets focus on upselling our customers on P-series.'
posted by pwnguin at 1:46 PM on February 9


We use DB2 and some IBM hardware in my lab. They're very good at squeezing license money from us, and absolutely terrible at providing actual support for those products. IBM far outlived DEC, but at least the customer engineers from Digital actually knew something.
posted by wintermind at 1:51 PM on February 9


In The Soul of a New Machine as Tracy Kidder outlines the genesis of Data General, one of the data points is that IBM was selling a machine for $2 million + which cost less than $200K to manufacture. DG was formed to clone such machines and sell them at more realistic market prices. IBM sold against this practice with the classic FUD approach; Kidder recounts a scene where the IBM salesman walked into a meeting where the CTO had stacks of DG documentation about memory sizes, speeds, and such all broken down, and the IBM rep preserved the sale by saying "Do you want a vendor or a partner?"

This was in the late 1970's / early 1980's. So IBM has been gliding on this sort of bullshit for a long, long time. Hollerith and his generation may have been legitimate innovators who found new ways to automate grueling and boring tasks, but their first and biggest customers were governments and they quickly mastered the gaming of procurement systems and only very occasionally looked back.
posted by localroger at 1:52 PM on February 9 [1 favorite]


As it happens, the intersection of embedded and super high performance is gaming consoles, and Power is absolutely killing it there, being featured in all three major consoles.

They did during the last generation, but not anymore. Both the Xbox One and PS4 have AMD CPUs (whose killer feature is a powerful integrated GPU, something that IBM probably can't match.) The Wii U still users PowerPC, but it's selling terribly.
posted by cosmic.osmo at 2:04 PM on February 9


I'll bite.

PS4: 4.2 million units
Xbox one: 3.9
Wii u: 5.86
posted by sieve a bull at 2:17 PM on February 9 [2 favorites]


I'm puzzled by the idea that clusterfuck projects only happen when The Government is the customer.

Of course they don't. But only the government gets to keep on keepin' on as if nothing had happened. Private companies that experience clusterfucks have serious negative consequences. People get shitcanned. Companies cease to exist. Remember the Edsel?
posted by valkyryn at 2:35 PM on February 9 [1 favorite]


Remember the Edsel?
Actually, I don't. I remember the K Car, though.
posted by thelonius at 2:36 PM on February 9


WiiU has been out for well over a year. PS4 and Xbox One have been out for a couple months. Nintendo themselves released a statement a few weeks ago admitting the WiiU was a failure.
posted by sophist at 2:37 PM on February 9 [8 favorites]


Private companies that experience clusterfucks have serious negative consequences. People get shitcanned. Companies cease to exist.

Remember how Coke's entire C-level suite got shown the door when they pulled the plug on a failed billion dollar implementation of SAP? Or how Hershey's execs were fired for nearly bankrupting the company by screwing up their logistics system just before the Christmas season that makes up the majority of their sales? Or RSA's head, Art Coviello, got publically fired for the security breach that exposed every extant SecurId customer to impersonation, or for selling out the entire company to the NSA for $10MM?

Yeah, me neither.
posted by fatbird at 2:56 PM on February 9 [18 favorites]


Yeah, me neither.

I'm not having a conversation about whether or not C-level execs are adequately penalized for failure. I mean, you can deliberately fail to take my point if you want to, but the fact of the matter is that the annals of business history are littered with the wreckage of companies--and employees--that screwed up one time too many.
posted by valkyryn at 3:03 PM on February 9 [2 favorites]


It's a fair point, valkryn, and I do take it. But I'm as annoyed as maxwelton by the received wisdom that private industry is somehow more accountable than gov't, because capitalism. As with failure to punish CEOs, we can also demonstrate administrations getting turfed out in the next election, political careers derailed, and (in Canada, at least) whole parties destroyed over massive boondoggles, corrupion, and costly failures. It's facile to point the finger in either direction.
posted by fatbird at 3:09 PM on February 9 [7 favorites]


I don't know how much I can spill here without getting into trouble with my job, but I am so glad this series of articles exists. I can say from personal experience in my day-to-day job duties that IBM is the worst. Just absolutely the worst. We have so-called server administrators who don't know how to do even the most basic things in Windows, such as renaming a file, or determining what file type an item is. Some of the people can't even listen to a basic description of the problem and form an understanding of what you're trying to get them to do. I used to be able to go to our server team to get help, but now I just do it myself, or find some way to work around them.

There are a few really awesome people working for my company under the aegis of IBM as well -- I had a good conversation with one of those folks yesterday -- but they fear losing their jobs as IBM's goal just seems to be hiring the requisite number of bodies as cheaply as possible in order to fulfill the nominal requirements of the contract.
posted by spacewaitress at 3:15 PM on February 9 [2 favorites]


Clarification: I don't work for IBM. I work for a large company that has engaged IBM to do many IT functions that used to be done in-house.
posted by spacewaitress at 3:16 PM on February 9


If IBM's future is in the services, they're structured the wrong way for it. A public company is not a good way to organize a high-skill, high-margin professional services company. The pressure to hit quarterly earnings numbers simply isn't consistent with the way you manage talent development and retention through the business cycle at a top-of-market consulting firm. Eventually you either turn into a low-margin offshorer/body shop like Accenture, or melt down completely like BearingPoint; sounds like IBM is already pretty far down that spiral.

Global Business Services would be a lot higher-value in the long run if it got spun back off as a partnership, or sold to one. At the rate things are going it'll probably get sold to one of the big Indian outsourcing shops, which would probably also be an improvement in terms of management effectiveness.
posted by strangely stunted trees at 3:17 PM on February 9 [1 favorite]


spacewaitress: "Clarification: I don't work for IBM. I work for a large company that has engaged IBM to do many IT functions that used to be done in-house."

I'm so sorry for you and your co-workers.
posted by octothorpe at 3:18 PM on February 9


I'm as annoyed as maxwelton by the received wisdom that private industry is somehow more accountable than gov't, because capitalism.

I'm not saying precisely that. I'm saying that the arcane edifice which is the government (and especially federal) procurement system has no obvious counterpart in the private sector. It's really bizarre, deeply counter-intuitive, and the reason that the government spends an order of magnitude more on any given item than private businesses do. Indeed, the procurement system is so byzantine that it's its own legal practice area. That system is precisely why companies like IBM and CGI Federal exist despite objectively abysmal performance.

Any company that tried to deal with private clients the way government contractors deal with the government would rapidly find itself without any clients. But the government exists, so that's not a problem for IBM.
posted by valkyryn at 3:19 PM on February 9 [2 favorites]


That system is precisely why companies like IBM and CGI Federal exist despite objectively abysmal performance.

Less than 3% of IBM's revenue is from the federal government. State and local is probably less than that. It's hardly the decisive factor in keeping their lights on.
posted by strangely stunted trees at 3:27 PM on February 9 [3 favorites]


cosmic.osmo: "They did during the last generation, but not anymore. Both the Xbox One and PS4 have AMD CPUs (whose killer feature is a powerful integrated GPU, something that IBM probably can't match.) The Wii U still users PowerPC, but it's selling terribly."

Whoops. Guess I wasn't paying attention close enough to the page of major PPC consoles I was looking at, and assumed I was looking at the page for this gen's consoles. Good times, IBM. Good times.
posted by pwnguin at 3:30 PM on February 9


IBM sold their federal division in 1994.
posted by octothorpe at 3:31 PM on February 9 [2 favorites]


IBM is pivoting (or trying to pivot) to a totally new market and new products/services here.

There is more to this, even than you think. I work/worked for IBM (until last week, when I was made redundant; no hard feelings I got a nice payout). Employees are well aware of the Cringely articles, and whilst he gets some things right, there are some really critical, big-picture things he gets wrong - and I'm in a decent position to judge because of my former role.

For the purposes of these articles, it's helpful to think of IBM in three phases - Pre-Gerstner, Gerstner, and Post-Gerstner. This is not to say that it's the best way to think about the company, full stop - but in this context it's helpful.

Old IBM was a - very, mindblowingly successful and profitable - hardware company. Then the nineties happened, and IBM got genuine competition to its products, and the profit disappeared. Most ex-monopoly companies start going bankrupt at this point, and most people have no, no idea how close IBM got. It got so close. The fact it didn't functionally disappear like Kodak etc is so unusual, and it came down to Lou Gerstner.

Gerstner, unlike most IBMers who were heavily invested in the old model and incapable of envisioning change, knew the company needed to change its business model, fundamentally - and it needed to convince the market, its clients, shareholders etc it could do it.

He recognised the growth of outsourcing in general, but services in particular and poured the colossal might and resources of the company intro transforming IBM not just into a services company, but into the services company, with all the prestige (and price, and profit) it commanded in hardware. When I say outsourcing, I'm not just referring to "sending jobs to India" or whatever, but to IT outsourcing; companies paying someone else to run their entire IT infrastructure. These contracts are incredibly lucrative, profitable, large, and best of all they typically run for years and the costs to change providers are prohibitive.

Gerstner - against all the odds, I might add - succeeded in this. IBM, though still firmly embedded in the consumer mind as a hardware company, became one of the blue chip services/IT consulting companies with a very large market share. Given that a) IBM was not known as a services company, and b) did not have a "unique" offering in this space (the way google did/does for search, for example), this is quite amazing. He returned the company not just to profitable but to very profitable. This started in the nineties and continued, roughly speaking until around the GFC ish.

And then the market changed again. What happened, partly as a result of the GFC, partly because of changes in technology is that the demand for IT services, IT consulting, and those huge outsourcing contracts started to drop, and then the drop sped up.

There were/are a few reasons for this. In a shaky, post-GFC environment companies definitely didn't want to get locked in to 5-8 year long contracts to the tune of millions, sometimes billions, of dollars. They started getting much choosier about where to spend their money, and how. As their returns started to drop, they started expecting better returns on their own contracts. Additionally, the "commodification" of the hardware market continued to move up the sophistication chain, and it wasn't just PCs now, it was servers, especially cheap blade servers (Bye Sun! Uh oh, HP....). What this meant was that a) servers got cheaper, b) it became much, much more feasible to run IT services in-house, and c) it was also a lot easier to migrate from platforms as one blade server is much like another.

The advent of Cloud Computing exacerbated all this - especially when players like Amazon and Google entered the market. These guys are comfortable in the cloud space with razor-thin margins because they sell such huge volume, and for big guys like IBM and HP - who need a healthy profit on each sale and whose shareholders demand a healthy profit - it was very, very hard to compete.

So, the model IBM had been hugely successful with for the last 15-20 years needed to change, again. That change is what we're seeing the company go through now.

The change has two parts: 1) Changing to a software company, and 2) Penetration in the developing world.

Regarding change 1), so if IBM needs to be profitable, and hardware on the whole wasn't profitable anymore, and now services is losing its profit, what is still really profitable? The answer is software. So IBM expanded its own software development, capabilities and products, but also - and very visibly - it started buying up a shitload of software companies, especially in the analytics and "big data" space. It is gobbling up more of these companies every single day, because that's where its leaders see the future profit, and frankly I see little reason to disbelieve them.

Software has great margins, but unlike Services and consulting, requires very little headcount. To put in perspective: When I started working at IBM in Australia, we had around 15000 employees. Microsoft Australia has around 550. So, needless to say, transforming into a software company, as less services contracts go around means losing a lot of employees, both because there is no need for them, and also because they are not software people.

2) The developing world. IBM less of its revenue from the US, and more from the developing world (especially China) than you think. This is not an accident, the company recognised that this was the next frontier for IT revenue in the 2000s and invested accordingly. They were a bit late to the game when it came to Africa, concentrating on Latin America and Asia, but they're trying to make up for it now, off the back of a few massive contracts like Airtel. This also means jobs going in the US and other western countries, to hire dozens more people in the developing world. If you are an IBMer, there are lots of opportunities in the developing world. I did 5 weeks in Kenya last year and came away with several guaranteed job offers - they cannot hire people fast enough. I believe that this transition is a smart move, on the whole.

Something interesting to note about IBM in Africa and the developing world in general: The company has a strict policy of "no bribes" - and they really mean it (at least, I can vouch for this in Africa, I have heard it's the same in China). You can imagine the challenge this means when it comes to winning contracts (in govt especially) in places like Nigeria and Kenya. I can tell you IBM's competitors definitely do not all have or enforce this policy. But the company is still winning work, so they're doing something right.

So, what you're seeing is IBM: Once a hardware company, then a services company, and now a company making the transition to a software company.

Cringely's critique seems to be "IBM made a lot of money being a certain type of company in the past, why can't it just go back to being that company?". It's naive, and silly, and focused both on a type of client that doesn't exist any more and very US-centric - something increasingly less relevant for the company. The company would go bankrupt trying to be the kind of company it once was, without a doubt. You can criticise a lot about IBM (myself and my former co-workers do frequently!). However, you can't criticise it for being blind - the experience of nearly going out of business from being on top has shaped the company and had an indelible effect on its leaders; they are determined it won't happen again. Ensuring that it doesn't is a slow business in an organisation the size of IBM, and it can be painful, but they are turning the jumbo around.

Miscellany:

His criticism of the company as caring only about shareholders and the market is both true to an extent, but also redundant. Senior leadership has immense pressure and incentive to do this. They need to do it because IBM needs that shareholder confidence (and money to a lesser degree); look at competitors who don't have it: Sun, HP, Oracle slowly but surely. It's not pretty.

When you are losing money as your business model changes, how do you secure that confidence? Through moves designed to maximise EPS and growth in profit, even if revenue drops. It's painful, but I don't really see any other way of doing it.

Also, leadership gets a lot of money for doing it that way - something I feel dogged Palmisano as General Manager and was one of the reasons I didn't rate him. Rometty is a very different kind of GM, and if anyone can make this transition it's likely to be her - she is smart, is not afraid of change etc.

On selling x servers to Lenovo: A very good move. Seems a lot of people don't understand this is just one portion of IBM's hardware division - the suckiest, smallest margin, highest competition portion. It's total commodity business, something IBM has always been bad with, and very very low profit (and lowish market share, well, low for IBM).

Mainframes and power servers - the kind of servers you need to run... big data and analytics software, have not been sold, won't be sold for a while at least, and remain profitable for the company. IBM can't compete in blade servers, selling it was a good move, and I would expect other sales in other, non-software areas to come (not inside info, just my feeling).
posted by smoke at 3:46 PM on February 9 [73 favorites]


An analogy: Imagine if Dominos said, "We're getting out of pizzas. We're starting to lose money on it, despite the brand, the popularity etc. We're going to become grocery stores. We'll still sell pizzas, but we'll be selling them in the grocery stores. Not only are we going to become grocery stores, we're going to become one of the biggest, most successful grocery chains - not just in the US, but in the world - and we're gonna beat established grocery stores, and we're gonna make money and grow doing it."

IBM has, unbelievably done that once, and it's currently trying to do it again.
posted by smoke at 4:11 PM on February 9 [7 favorites]


Re: maintaining investor confidence and profit during giant shifts, the obvious example of why this is important is JC Penney.
posted by kiltedtaco at 4:17 PM on February 9 [2 favorites]


"Problems that an experienced person could fix in a few minutes are taking an army of folks an hour to fix."

Yup. I have personally experienced this. Now, I just avoid working with IBM altogether. I either fix it myself, or if it's a single-user issue and it depends on an IBM person to fix it (because the IBM people have more permissions to the servers than we do), I flat-out warn the user that their issue just may not get fixed, at all. (I still send a ticket to the appropriate group, knowing that they may fix it, or they may ignore it, or -- the most likely option -- they may do something to make the problem even worse).

It's actually surprising to me how well our IT organization has been able to route around the deadweight of IBM and still get stuff done. There are still enough people around with knowledge, skills, and willingness to work (some of them are former employees who were laid off by the Company and then re-hired by IBM as contractors). But what happens when there are just too many holes? What happens when there's nobody competent left to get things done? Will things just grind to a halt? Fall apart? What then?
posted by spacewaitress at 4:43 PM on February 9


I worked for IBM (in the US) for ten years, mostly doing operating systems programming. It was overall a net positive experience. I got to work on really interesting problems with some of the smartest engineers in the world, and I have many treasured memories, though I hope I never have to use Lotus Notes again.

My working environment was very pleasant and professional without being too uptight. I could come to work in shorts and a t-shirt (and often did). I can recall being yelled at only once. Harassment was not tolerated, as far as I could tell (disclaimer: straight white guy POV). I'd like to think the company has been progressive in some areas; for example benefits were extended to same-sex partners starting as far back as 2001.

I left a couple years ago. I had gotten stuck on a doomed project, and was getting passed over for promotions. There wasn't any immediate peril, but I felt I was getting more expendable the longer I went without climbing the ladder. My timing seems to have been good; there's been some serious bloodletting since then.

It was easy to find something that paid better -- I've never thought IBM was very competitive for engineering salaries, but the benefits seemed pretty good. I find it a little shocking that they've pulled this stunt with the 401(k) match, considering.

I didn't know Cringely spent so much energy covering IBM (he's even written a book on IBM to be published soon). Some of his criticisms ring true, especially with respect to the services business, but he's way off on other stuff.

Like smoke and some others have said, selling the x86 server line was just good business. That was not a matter of if but when.

I think the next couple of years will be make-or-break for the PowerPC business. I"ve seen some indications that they may be making inroads in big data. Perhaps they can carve out a niche there.
posted by scatter gather at 4:55 PM on February 9 [1 favorite]


Yeah, IBM did the AOL 401k thing last year. It's so transparently bad for the employees, especially in a time of layoffs, that I'm surprised they didn't find some more obfuscated way to do it. That's a lot of cash to lose if you get fired or quit near the end of the year.
posted by smackfu at 5:20 PM on February 9


Mainframes and power servers - the kind of servers you need to run... big data and analytics software, have not been sold

Is this a new development in big data computation? As far as I know, none of the big companies known for large scale analytics (Google, Amazon, etc) use mainframes and power servers. The sense I get from the "big data" community is also that large arrays of cheap computers are preferred.
posted by lenny70 at 6:27 PM on February 9 [3 favorites]


As far as I know, none of the big companies known for large scale analytics (Google, Amazon, etc) use mainframes and power servers. The sense I get from the "big data" community is also that large arrays of cheap computers are preferred.

I guess it depends on the nature of the datasets, and how easy it is to partition them or parallelize the work across the large arrays of cheap computers. For some cases it is easier to use a big mainframe. Example, for relational databases, partitioning the database itself can lead to a whole slew of issues regarding data access time, data update time etc.
posted by all the versus at 6:41 PM on February 9


The sense I get from the "big data" community is also that large arrays of cheap computers are preferred.

This is definitely the case with Google, and much of their early research (after search algorithms) was in how to make such networks deal gracefully with failed nodes, which they do not bother repairing but just leave in place as the work is routed around them. I would imagine Amazon is doing something similar.

On preview, all the versus is right that some problems can't be easily scaled this way, so there will always be some market for reliable big iron. However it will be a high dollar specialty when it's needed.
posted by localroger at 6:43 PM on February 9


In those six year, I never did figure out what IBM did for a living. They certainly weren't even slightly interested in developing software.

There is "IBM Software Labs." IBM bought a tech company in my city about 10 years ago that does XML forms. They grew the unit for a few years, but unfortunately in the last year or so there seem to have been layoffs.
posted by KokuRyu at 8:30 PM on February 9


lenny70: "Is this a new development in big data computation? As far as I know, none of the big companies known for large scale analytics (Google, Amazon, etc) use mainframes and power servers. The sense I get from the "big data" community is also that large arrays of cheap computers are preferred."

AFAICT, Big Data is more about scatter-gater techniques and data parallelism than pure number crunching. I'm not aware of any particular advantage the Power arch would have in retrieving or storing lots of data. And I guess Hitachi now owns the particular limb of IBM that dealt with storage.
posted by pwnguin at 12:51 AM on February 10


IBM sold their federal division in 1994.

Well color me surprised.

So, IBM surviving based on federal contracting revenues shot down as a theory. Done.
posted by valkyryn at 6:28 AM on February 10


smoke: "We're getting out of pizzas. We're starting to lose money on it, despite the brand, the popularity etc. We're going to become grocery stores. "

IBM wasn't losing money because people were buying fewer of their PCs and servers. It's because people were buying fewer of anybody's PCs and servers. It was an industry-wide shift.

If Dominoes sensed that the entire population was beginning to lose its taste for pizza, they'd almost certainly try to rebrand, or use their capital to buy another business to latch on to. Heck, this is exactly what the Big Tobacco companies have been doing since the 1980s.
posted by schmod at 7:30 AM on February 10


Like every other business in the U.S. they've sold off every concrete thing that ever made them interesting or unique to become a consultancy. It's a recipe for disaster all around.
posted by xammerboy at 11:10 AM on February 10


A pizza consultancy?... wait, I'm confused now....
posted by infini at 12:06 PM on February 10 [2 favorites]


The choice wasn't "keep doing something interesting or become a consultancy"

It was "Try to reinvent the tech side - if we fail we go away. Become a consultancy - if we fail we go away"

The odds on the first one are really really small. The odds on the second one are ok, not great.

Even a business as diverse as IBM in 1985 was built on a single digit number of home run creations that they discovered over the prior 75 years at completely unpredictable intervals. Each one of them was very low odds and there was lots of capital burned on doing that. It worked because you had the core business that was able to subsidize lots of pretty basic research that led to these home run innovations. But there ends up being a massive dispersion in returns. Probably 95% of the R&D spend IBM has done has basically 0% ROI while the remaining 5% was probably 1000x or 10000x.

By the time Gerstner came to run the place the economics of that just didn't work anymore. Or at least you could see on the horizon what was going to happen. He had to decide how he wanted to bet the company. If you bet it on innovation its tremendous risk reward - and basically unpredictable. If you bet on services its much lower risk reward and the downside is much less negative. Unlike R&D every dollar spent on building a services biz gives you a much tighter distribution of returns - with the cumulative return being lower than it would be if had infinite amounts of capital to invest either in R&D- but that's the tradeoff. But you have a finite amount of money you can spend. IBM could have spent 5 million USD and created google. It could have spent 50 billion USD and gotten nothing

You've seen this happen to pretty much every big Tech or Industrial Company if you have a long enough time frame.

You can actually make really good money betting on the failure of midsize tech companies built on one product that has commoditized and they are using the cash they've got to try to hit another homerun.

At some point it'll happen to Google.
posted by JPD at 1:32 PM on February 10 [1 favorite]


At some point it'll happen to Google.

LOL, no. Google has been incredibly agile in seeing which way the wind blows and understanding how its bread is buttered. They're shooting for another home-run now, while times are good. This used to be Microsoft's strategy before Bill retired - see where people are starting to make money, and then steal their business. Don't wait until you're in a hole to start climbing.

This is what infuriates me about Apple at the moment - they have 160bln on hand, and they're not re-investing it. They should be broadening the fuck out of their horizons with a significant chunk of that money (and sitting on the rest for a rainy day - 5bln in cash on hand kept it out of the hands of Sun and Bankruptcy Court in the '90s. Keep ssssayyy - 25bln on hand.)

Apple's R&D culture and unique market position could steal both Intel's and ARM's lunch money with something new and cool, blow up the networking world with something as powerful as an Aruba but as easy as an Airport, make Mach-based virtual systems designed as "appliance images" - dumbed way down for whatever specialized app you need it to run, designed to be deployed into whatever cloud service you like from a Mac with a few clicks controlling a vast infrastructure, I mean the list is endless.

Worse, it's socially irresponsible. Money spent on share buybacks or sitting in cash is money not used to invest in technology and more important, people - investors and workers both are losing out.
posted by Slap*Happy at 1:48 PM on February 10


LOL. It will. Look at a list of big tech companies 40 years ago. The tech companies that survived either don't do innovative tech, are currently facing existential crises, or no longer exist.

Of course google is shooting for home runs. As they should be. But anyones ability to hit them is much lower than they think at the time.

Essentially Googles ability to hit a home run isn't substantially different from your ability. The difference is they get a whole lot more tosses to get it done.
posted by JPD at 1:54 PM on February 10


At some point it'll happen to Google.

Probably but IBM is 102 years old so it might happen to Google long past my lifetime.
posted by octothorpe at 2:05 PM on February 10


That's true. Though remember - IBM and HP are both huge outliers.

Lots of iconic tech companies only lasted 30-40 years. Lots of really important smaller tech companies lasted less than 20.

Its part of why tech companies share prices react so viciously to seemingly small variance in results.
posted by JPD at 2:11 PM on February 10 [1 favorite]


I'm puzzled by the idea that clusterfuck projects only happen when The Government is the customer.

Yeah, the idea that private enterprise does everything Right while government does everything Wrong is a myth. Governments have a duty to disclose stuff like that, and it's going to be reported (in, interestingly, privately-owned newspapers etc.), whereas private companies don't have to tell anybody shit.

(That's a simplistic take on it, sure, but that doesn't make it any less correct.)
posted by turbid dahlia at 4:56 PM on February 10 [1 favorite]


Lots of iconic tech companies only lasted 30-40 years. Lots of really important smaller tech companies lasted less than 20.

Wang will rise again.
posted by entropicamericana at 4:59 PM on February 10 [2 favorites]


The amusing thing is when private companies are ripping off the government somehow only the government is to blame in the transaction.
posted by octothorpe at 5:29 PM on February 10 [4 favorites]


This article on the ten first software companies (actually the first seven plus Nintendo, Microsoft and Apple) is an interesting read - of the ten companies profiled, only one went out of business.

And if you look at Atari, they managed to define new markets and fuck them up irretrievably no fewer than five times (Arcade consoles, home consoles, 8 bit computers, 16 bit computers, palmtops) - would an Atari headed by an Ellison or a Jobs have gone under, with the tech portfolio it had in the early '90s?

A culture of constant innovation, contempt for industry consensus and a ready willingness to cannibalize your own best sellers for the new thing can take a tech company faaaaaaar. It does need someone who knows how to balance a fucking checkbook, Commodore, looking at you there, and Google has it all going on.

I'm kind of afraid Apple doesn't. Cook doesn't know how to pick winners. The single processor Mac Pro was a really dumb move - beautiful design that needed to accommodate another processor.

The new iWork apps are something of a disaster, MacOS has been lounging around on its laurels for way too long, and I'm sorry, but the brushed aluminum thing is really old, where the hell's my carbon fiber Macbook with dual octo-core ARM chips and 16 hour battery?
posted by Slap*Happy at 6:40 PM on February 10


Lets go through that list as it very much makes my point.

1) CTC - Now Part of IBM
2) IBM - Services biz these days
3) Computer Usage Company - Ch.7 Liquidation 32 years after foundation
4) CSC - Services Provider to Governments
5) MSC Software - was sold for 386 Mil USD in 09. Spent 400 mil buying other software companies from 94-09
6) Applied Data Systems - Lives off serving the perpetually shrinking mainframe business. If this grows it grows through M&A. Certainly aren't investing in R&D
7) Cincom - Appears to be a real software business
8) MSFT
9) AAPL - which don't forget was nearly consigned to the dustbin of history itself before it pulled off a very low probability come back.

So two of the companies on that list are still software businesses doing what they set out to do at the outset, one hit a hail mary - and the rest have abandoned the biz, are dead, or are dying.
posted by JPD at 6:01 AM on February 11


So if you think MSFT has basically transitioned to its conglomerate days - that's one out of 9 in a biased sample that is still truly a software business.
posted by JPD at 6:05 AM on February 11


Governments have a duty to disclose stuff like that, and it's going to be reported (in, interestingly, privately-owned newspapers etc.), whereas private companies don't have to tell anybody shit.

In practice, governments try to conceal this stuff as much as possible and are successful to a surprising degree. Even if the information is theoretically out there, most people simply don't give enough of a rip for it to matter. And again, a private company that fails sufficiently hard will go away, whereas governments just keep on keepin' on.
posted by valkyryn at 11:53 AM on February 11


entropicamericana: “Wang will rise again.”
Not long ago I was stunned to find out that Unisys is still around. Apparently so is one of the other titans of '80s IT in Atlanta: American Software, a company so dumb they had a thousand programmers in their headquarters and not one coffee machine.

Sadly Control Data only still walks the earth as Ceridian.
posted by ob1quixote at 9:33 PM on February 11


with dual octo-core ARM chips and 16 hour battery

ARM has certainly captured a sort of mythology.

If you care about single thread performance, especially for branchy code, ARM is not there and will not be there anytime soon, and when they get there they will pretty much design braniac cores that use just as much power as Intel.

There's just no magic coming. The way that so many people who have not looked at the actual reality feels like a sad replay of the PowerPC debacle ("It's RISC, of course it will be faster" -> I can't find the actual Motorola ads in the 601/603 timeframe, but this gets to the nut of it.

See also: We find that ARM and x86 processors are simply engineering
design points optimized for different levels of performance, and
there is nothing fundamentally more energy efficient in one ISA
class or the other. The ISA being RISC or CISC seems irrelevant.

posted by rr at 10:32 PM on February 11


IBM layoffs strike first in India; workers describe cuts as 'slaughter' and 'massive'.
posted by oceanjesse at 4:36 AM on February 12


If you care about single thread performance, especially for branchy code...

And we don't, otherwise we'd be running POWER or Fujitsu (Oh, look, RISC is still out there at the high end! Despite an enormous R&D budget from a near monopoly position, being hobbled by an interpolation layer for an ancient ISA allows much smaller competitors to keep pace. Whodathunkit?)

We do care about power consumption and price/performance when it comes to mobile workstations like laptops. Most applications are getting much better about paralelizing their code with modern dev tools, too.

That paper trips a lot of baloney warnings - and it runs contrary to what the industry is actually doing. ARM's price/performance and performance/power-efficiency keeps trending up, but Atom seems to have trouble keeping pace. No-one is really interested in adopting x86 in the mobile arena apart from a few niche players, and that folks are even talking about something other than Intel in commodity servers is a huge reversal in fortune, and driven by a projected trend rather than a snapshot of where we currently are.
posted by Slap*Happy at 5:09 AM on February 12


The people I feel sorriest for in all this are the people who look for certainty and consistency in their lives. They're the kind of people who choose the big corporation over the small startup, who shun the risk of change for the risk of stagnation.

It used to be that this kind of constant life was possible, at many levels of society, for all sorts of professions. It's been less and less true for at least the past three decades. Some people at IBM probably expected to be there cradle to grave.
posted by ZeusHumms at 12:05 PM on February 12 [2 favorites]


ob1quixote - you might enjoy the list of the Nifty 50
posted by JPD at 12:46 PM on February 12


ZeusHumms: "The people I feel sorriest for in all this are the people who look for certainty and consistency in their lives. They're the kind of people who choose the big corporation over the small startup, who shun the risk of change for the risk of stagnation.

It used to be that this kind of constant life was possible, at many levels of society, for all sorts of professions. It's been less and less true for at least the past three decades. Some people at IBM probably expected to be there cradle to grave.
"

I didn't but my did sure did and was darn proud that his son had hooked up with a "good company" and would certainly rise through the ranks and then retire there. I never had the heart to tell him how far IBM had fallen and he passed away before I was let go so he never knew.

I live in Pittsburgh which was one of the centers of that cradle-to-grave expectation from a job. People here were identified by there employment with Westinghouse, Alcoa, USS, J&L, Mellon, PPG, etc and would have never thought of jumping from one to another and expected the same loyalty from the company. I'm sure that some of my older neighbors would be shocked by my resume which shows six jobs in the last fifteen years. That was just unheard of here.
posted by octothorpe at 3:01 PM on February 12 [2 favorites]


« Older Forty eight, uh-huh, count 'em, FORTY EIGHT Fela r...  |  Informative article and hilari... Newer »


This thread has been archived and is closed to new comments