Is HP going to disappear? According to this article its problems are increasing rather than being solved. And all the potential solutions are weak, at best, and contain little hope for solving the problems. Regardless of how one feels about Meg Whitman, yet another CEO is unlikely to be the silver bullet.
I’ll get to the commentary on the business part but this is a personal story, so let’s do that first. Bear with me.
This is a sad story for me personally. My first “real” job after graduating from bizschool was HP, back in the days when Bill and Dave personally ran the company and it was still small enough to be managed. Bill and Dave had perfected a formula that led to HP’s spectacular success and being one of the most admired companies in the world. And certainly one I was privileged to join and that I can honestly say I loved.
I only knew a bit about HP when I went to the job interview. All I really knew was they were in Palo Alto and after 8 years in frigid Boston Palo Alto sounded like paradise. I clicked with the interviewer and soon found myself on a plane to SFO, my second on-site interview. It was the longest flight I’d ever taken and arriving in the confusing SFO I finally found my rental car and got my first experience on the Bayshore trying to find Palo Alto, my first turn onto University Ave and then El Camino – a whirlwind. When I arrived at the old Rickey’s Hyatt House I was severely jet-lagged but also shocked. My winter clothes from Boston were suffocating and everything, even in the dark, was GREEN! The week before I almost got snowed in trying to return from a Xerox interview in Rochester and I’d left Boston with its usual interminable gloomy dirty snow winter. Where was I? – another world! exciting for 23YO.
The next morning I drove to the original corporate headquarters at 1501. California is weird, green in the winter, brown in the summer, so the area between Page Mill Rd and Stanford was all empty then, beautiful grassy hills, what a great place to buy a lot and have a nice house (little did I know about Santa Clara real estate). I even remember the song that seemed to be playing continuously but that I’d never heard, America’s Horse with No Name. I was on another planet. The interviews plodded on and I was unsure whether I’d passed but little did I know that HP was just recovering from a recession and would hire about anyone with a pulse in one of their rapid growth spurts. The HP35 calculator had just been released and had amazed the world and HP had become the place to be. I was sold, just give me the offer, any offer.
The offer did come and despite a hard push by Xerox that muddied my decision, soon my signature was on the offer and here I come, Palo Alto. A rush to finish my course work, thesis, hand over my existing job, and finally I awoke one morning at about 6am with the movers wanting to come ship my meager possessions to a place I’d only visited for two days. After just a few hours sleep from my last day in Boston we’d made no preparations for the month-long vacation drive to California and had to run around grabbing stuff and throwing it in the car before the movers packed it and we wouldn’t see it for a month. After that flurry I can still remember the last night in the apartment, sleeping on the floor in the sleeping bags for our camping trip. Then off in the Datsun 510 for the longest trip I’d ever taken.
Across the country, stopping along the way in beautiful spots finally we’re on I-80 and, then, unwisely, heading down El Camino all the way to Palo Alto (the only time I ever did that). A couple of days of apartment hunting (strange world: swimming pools everywhere, decks, Mediterranean architecture, and WARM, but now being June, also those beautiful green hills were now BROWN, WTF, is this a drought?). And then my first day rolls around. We had only one car but we’d picked an apartment on Arastradero so I thought walking would work fine. The HR guy made a joke, that scared me thoroughly, that I couldn’t have the job since my employee number would make me 100 in division 00 (corporate) and Bill had decreed that corporate would never exceed 100 people, my first exposure to how HP really worked (completely the opposite of the usual top-heavy corporation). Then the thrill of getting to my desk, one of a group of four in a “bullpen” of 100 or so people in a single room (not even partitions in those days, much less offices – yet another shock). In my first few days I saw people I’d known in Boston since brass rats were a dime a dozen at HP (even outnumbering Stanford). On Friday someone noticed me working at 3pm and asked whether I was going to the beer blast – the WHAT? Some division had exceeded its MBO goals and so could spend on a celebration: a live band, beer flowing, burgers cooking, outdoors – what is this place! – literally an extension of college. And then soon it was picnic time, yes, another ritual up at Little Basin, replacing the original picnics (long since outgrown) at the Elk club for the real old-timers. Seeing both coffee breaks go and later picnics go were symbols of the end of the golden age, long forgotten now in the huge conglomerate HP has become.
As the first week turned into the first month and then the first year I gradually got indoctrinated into the “HP Way”, a wonderful set of rules, one of the first of “corporate cultures”. I’d already bumped into Bill when I grabbed my coffee in 3U instead of my bullpen in 3L (twice daily coffee breaks. a socially unifying exercise, eventually abolished as HP began to lose its way) were a ritual, one all the old-timers loved since it meant everyone could met everyone else. The informality, unusual in that era, no ties by anyway and even Bill (just upstairs) didn’t have an office, no assigned parking places, none of the hierarchical trappings common in corporate America at that time.
But the big part was the innovation kept flowing out, when 10% of sales was automatically directed to the “labs” (not engineering, the “lab”, every division had a lab even though there was one “HP Labs” which was a very exclusive place). And all that innovation led to great products with nice high profit margins, given the conventional Pi x cost (actually ranging from about 3.0-3.5, but sometimes 3.14159 just for fun, yes, marketing was that geeky). The stock had been a skyrocket, profit sharing pumped up our paychecks, everyone was in the stock purchase plan. At that time HP was very reluctant to hire anyone who had prior experience, largely because they wanted to combine the enthusiasm of youth, freshly trained in elite colleges (mostly male, mostly white (in skin and shirts), mostly U.S.) to then be molded by the “HP Way”. In many ways it was a cult and the new initiates either bought into the system or left. Despite the rumored version of “lifetime employment” you could get fired at HP for performance but otherwise you had a lifetime job and you were part of a family. If what you were working on got shut down you went on a list that must be hired from before outsiders were hired. And you were expected to adapt to a changing world and constantly re-educate yourself and steadily learn new skills (performance evaluations were strongly based on comparing to previously set goals, MBO was not just a buzzword, it was the “way”, both individually and through every organizational level (and not many of those).
What made HP work, other than having the best products so they could get away with those high margins, was the incessant innovation, driven by having the smallest possible almost completely independent business units. HP had just acquired its sales force (previously it had used dealers and partners and independents) so it had a tradition that was more wholesale than retail. Divisions had to sell the sales force first so despite being called a technology company HP was in fact intensely customer driven.
But it was who those customers were that made it work. HP’s business, when I started was almost entirely instruments, mostly electronic (and some analytical) measuring devices. They were sold to electrical engineers. And the engineers in a lab were electrical engineers. So HP’s market research was simple, known as the “next bench syndrome” – when you had an idea or were working out details of the product you just took your ideas to someone at another bench (I had a desk, as a software type, but the “real engineers” were EEs who had lab benches, cluttering with all the tools of their trade.) If the guy at the next bench thought your idea sucked, back to work, if they loved it, ship it.
In those days career path (at least for future managers) at HP was simple. You started at a bench and got a product out. As a new engineer when your first product went MR (manufacturing release) you went with it and had a new job title, manufacturing engineer. You rolled up your sleeves and actually built the sucker. And yes, HP built stuff and almost entirely in the U.S. In fact, every day coming into work I walked through a sheet metal shop when machines loudly banging out chassis and panels. Once we software types complained about the noise of stamping presses and actually had a hearing with Bill whereupon Bill explained he wanted every employee to know “we make things here.” End of new floor plans, the stamping presses stayed. After a stint in manufacturing it was usually back to the lab and at some point it was out into the field, either selling or in some technical support role. The HP Way incorporated the concept of the “triad”, engineering, manufacturing, and sales. You could stay in engineering or manufacturing or sales forever, but if you wanted to advance you had to work in all three and actually know something,first-hand, about how these functional areas worked. After succeeding in all three areas and getting a couple of products out, given you wanted to, you began a management path, after many years (some superstars cut it shorter) and mastering the HP Way. By the time you made division manager you KNEW the business.
It was a great formula for a company with thousands of high tech products sold in relatively small quantities with features, not price, selling the products. And the profits rolled in. When there were slack times people found things to do (often reorganizing inventory, tearing down and rebuilding a manufacturing area, doing a few more advanced products) – no layoffs. HP pioneered the idea of across-the-board temporary pay cuts (and reduced hours, more time to play in all that California sunshine). Usually increased profits put most of the pay cut back in your pocket via profit sharing. And all that stock you’d bought every quarter, your wealth steadily rose.
It was beautiful. It worked.
Until it didn’t.
In my sense of personal history what broke HP was entering the computer business. While I wasn’t directly involved I talked to many who were. Dave was hot to do it (the growth potential) and Bill was reluctant (low margins, complex new business where HP was a nobody). But instruments themselves were gaining “smarts”, the beginnings of the pervasive digital technology was being merged with the old analog stuff so instruments computed as well as measured. So sooner or later someone would want to build a computer instead of putting smarts in every instrument. In fact, from the folklore I picked up, that was the spin that got over Bill’s resistance. A clever young marketing guy, Ed McCracken (later head of the first computer division), sold the first computer as an “instrument controller” to sneak it by Bill. The ice had been broken. DEC had really fired the first shot, creating a new market niche for minicomputers, an area not dominated by IBM. So HP rode the coattails of that and sold could avoid the might of IBM’s marketing by doing what HP was good at, building stuff for engineers. But Ed (and others, not completely sure whose idea it was) had heard of Dartmouth Basic, and soon the HP2000 was born, a fantastic product. A low cost computer that could provide simple time-sharing to 16 users for a cost per user about the same as early PCs.
Note: It was about this time that Steve and Steve working in APD, in minor roles. Divisions did an annual event, sorta like a science fair, where all the protocols were shown to a visiting horde of managers. Steve and Steve were in love with a “calculator” that was actually a very early personal computer. All it could was play blackjack, and according to rumor, Bill’s comment was “we build serious stuff here” and Steve and Steve decided some people wanted toys and now Apple could buy HP with petty cash.
Then came the HP3000. It was an incredibly ambitious effort. Again it had a spin, it wouldn’t go head to head with IBM because it focused on a niche, interestingly opened by my other job offer, Xerox, for “scientific computing”. Lots of stories by people more involved than me have been written about the HP3000. Basically it was a failure. As usual the hardware was ok, but the software was a mess. Given that at that time software engineers were definitely second-class citizens (while there might not have been a lot of hierarchy at HP there was definitely a pecking order in job status and software was viewed as pseudo-engineering and people doing it were treated as less-than-engineers). That attitude, plus too many hardware guys pretending they could do software because it was trivial, but way too small a budget for the OS led to a miserable machine.
The scientific computing market was a bubble so before the 3000 was released it was already refocused on business computing. Now in the first half of my career I wasn’t in a lab but actually in the corporate computer center, known as BAEDP, naturally based on an IBM mainframe. With the refocus on the business market, John Young decided all the people sitting below him in BAEDP should play a role; we were the “next bench” since we did all that boring COBOL stuff for payroll, accounting, and manufacturing. This role was largely based on the tremendous success of COMSYS, an internal communication system (when the Internet was still ARPANet and largely a research project). Those high margins meant HP could buy its own minicomputers cheap. And telephone lines were expensive. So some clever people, led by Rich Nielson, built an electronic communication network that minimized expensive phone charges by using lots of processing power. Every office had a minicomputer (unaffordable to any other enterprise else paying high IBM’s, or even DEC’s, retail prices) and the IBM mainframe doing what today a Cisco router would do. As a consequence nearly-real-time order processing was possible and HEART (the order processing system, my first assignment) was born. Tagging along was the first global email system that connected all offices everywhere around the world.
So someone would get the first internal HP3000, outside the development labs, and I (as the most geeky of the “Cobol” guys) got it, the old “pizza oven” machine. It was awful, just barely usable. Being used to the power of an IBM 370/155 the HP3000 was an abacus. But gradually we began to build software, initially only for internal use, to try to turn it into a business machine. I got my career break as HP was desperately trying to sell the 3000 to business market and had no clue about IBM technology, or especially terminology, so as the manager of “systems programming” (today’s sysadmin) sales recruited me to attend customer sessions to translate.
Skipping a lot of detail my involvement in that effort eventually led to me going to do “real work” in one of the divisions, actually now a manager in the labs (only my brass rat saved me from the skepticism of the “real engineers” I managed). And that’s when I began to see the marketing weakness in this whole plan.
Bill hated big organizational entities. But a computer system is a large project. So the solution was a bunch of divisions building bits of the system. The old HP Way where each division was separate and its own profit center (really more than the usual notion of profit center, really its own little company under a single corporate name). There was little interest in and experience in coordinating across divisions. One of my first products in my division was shown in a trade show in the DEC booth (who supported the product) and not in the HP booth (who didn’t support the product), not exactly the best message to put before customers. With every division doing its own thing, guess how well the “system” was integrated.
In that era, before global video, another HP tradition was NPT (New Product Tour). Rather than every division going out to sales office, multiple product launches were packaged in the “tour”. People from the divisions, including the lab, fanned out to train all the sales offices, usually in a single week, on all the new products. Meanwhile the division team met with potential customers and did the sales pitch themselves, with the sales reps watching and learning how to make the pitch. In one NPT I participated in, HP’s first laser printer was introduced. It was a beast and had a lot of issues, primarily because others owned most of the basic patents and HP had to use inferior basic printing technology (my old potential boss at Xerox was probably laughing at me that I didn’t take his offer instead of upstart HP). The sales force hated the product and a few of the marketing people at the Boise division, heroically, sale by sale, got the product into the hands of customers. Once sales realized it was possible to sell the beast and the commissions were huge, it began to take off. Who knew that this product’s descendants as well as the thinkjet printhead (a true HP invention, originally for calculators) would become the powerhouse that would eventually fuel most of HP’s profits.
But the fragmented engineering and sales effort was taking its toll. I remember a briefing by Dave where he vigorously denounced what we were all doing. As part of HP’s sales push various senior executives were working their way onto boards of large potential customers, a favorite tactic of IBM. Dave was on the GM board. At one point he asked GM why they didn’t buy more HP computers. The answer he got, which no one short of being on the board could have gotten, was that HP divisions and the divided per-product salesforce competed more with each other than they did with DEC or IBM. Dave decided to fix that. First the sales force was reorganized around collections of products aimed at a particular task and then also around segments of the market. That was a revolution but not without huge pain and suffering and some vicious infighting.
But then what to do about all those independent divisions. That was the really tough nut to crack. Decades of the HP Way and independence of divisions had to go. Talk about a revolution and civil war. All this happened at the same time as one of the first really big startup booms, the PC. Hoards of HPers, dissatisfied with now having to be a drone in a corporate structure fueled many of those startups. The HP Way, inherently, was “small”, not the corporate bureaucracy needed for a huge and coordinated effort, so a startup felt a lot more like home than the now strange new HP.
So as far as I’m concerned that was the beginning of the end. Dismantling an ingrained corporate culture and reinventing a new is extremely hard. And HP never really accomplished it. And all this was happening at the same time as Bill and Dave were beginning to turn over the reins to a new generation. Lots of infighting for power thrown on top of a civil war completely destroyed the old culture without having built a new one. Meanwhile the classical career path (described earlier) was also a problem. In the old system every manager had been an engineer somewhere along the line. And if software engineers were second-class, marketing people were below that. And they resented it, especially as they were being given more power. So now another civil war broke out, crossing divisional lines because this war was now between job functions.
By the time I left HP a substantial chunk of the old-timers had left. Jimmy Treybig had shown all of us a new path. Again through folklore I heard Jimmy took the fall for the HP3000 debacle and was one day handed his walking papers by his previous subordinate, Ed McCracken, who got Jimmy’s job and then did the axe job. But Jimmy had hardly cleared out his desk before he got a call from a VC, “how would you like to start a company?” Later I met that VC (I now can’t recall his name), but learned the inside details that he’d been one of Dave’s finance guys and had left to be sorta like today’s hedge fund manager, to basically invest Bill and Dave and a few others money. So literally HP money funded the startup of Tandem. And Jimmy got rich. And the rest of us wanted to as well.
So HP became a kind of entrepreneurial grad school. After college go learn how business and management really work and when you got screwed in internal HP politics, go do a startup. The talent drain was gigantic. But also this meant a lot of newhires at HP, and the old get-the-young first-jobbers was gone, HP needed instant marketing organization so they hired a lot from strange places like P&G and EDS.
[Sidenote: One amazing irony is that Tandem, DEC, EDS all ended up back inside HP, but well after they’d all begun their own declines]
So the combination of virtually replacing all levels with outsiders from entirely different backgrounds, destroying the old HP way, losing so much talent, and management transition at the top doomed HP. But the real thing I think also helped kill it became the aversion to innovation. Since the “lab” owned innovation that civil war between engineering and marketing (the new hires) also led to de-emphasizing innovation (since that turf was owned by only one side in the feud) and HP switched from the leader to the copier. And unlike Microsoft they weren’t very good at it. Mushy undistinguished all-things-to-all-people products became the norm that no customer really liked and had to be pushed hard by an increasingly aggressive sales force to keep up the quarterly results. Meeting the numbers took on new meaning (more like most modern companies today, faked results) and a lot of stupid stuff was done to get numbers.
By the time Carli showed up (the ultimate outsider) and declared the HP Way dead, most of the damage had already been done. Carli may have driven the silver stake through HP’s heart and started the end of the end but the decline was a long process already underway. Carli’s lack of insight led to creating new messes instead of fixing any of the old ones.
Now HP has lost its way. Its disastrous history of CEO-de-jour just piled insult on top of injury. There is only so much damage a once-great organization can take before it’s just walking dead.
Now Meg gets to try to turn this around. As armchair quarterback my feeling is HP is beyond a few cures. It has cancer and has to have some major surgery. It’s too big and cumbersome (sorta like CitiCorp) to ever be managed. Its mission is too confused to ever become crisp and agile. And clearly now the famous profits are disappearing. So if HP shares are ever going to be worth anytime in the future (fortunately the ex got all of mine so I don’t care) I think really drastic change is needed. That probably means breaking the place up, a lot. The Compaq acquisition was a disaster only an AT&T opportunist like Carli could love and was the fatal blow and must be undone. So I think, unfortunately, Meg has to do what Carli did at AT&T, package (badly, but with cooked appeal) the bits and pieces and chuck them and mostly return the money to the shareholders. Shrink HP to 1/10th of its current size and see if it possible to restore innovation in whatever product lines are retained.
But CEOs are usually paid based on size so it’s unlikely Meg will do this. She’ll lay off tons more people, just further reducing already non-existent morale, and play accounting games and HP will continue to bleed. Interestingly, what HP needs is some of Mittwit’s Bain approach and a little of Boone Pickens. Create discrete pieces, probably loaded with debt and unlikely to survive, sell for more than the whole. Maybe HP will start the new wave (of a constantly recurring cycle in business) of the old disintegration rather than mergers (buying more businesses into the already dysfunctional culture will be a disaster). HP and CitiCorp need to be busted up. Maybe Obama could have, had he not listened to Geithner, saved CitiCorp through busting it up, but HP is on its own, so it’s Meg’s problem.
So a great company turned itself into a terrible company through a whole set of one step after another and now is a Gordian knot. Who will be the Alexander who just hacks the knot to pieces?
Disclaimer: Despite all the personal anecdotes I include in this post, I readily admit to only being a bit player in this drama, but at least I was close enough to the real players to have a fair amount of off-the-record knowledge of some of the events.
p.s. On the personal note, leaving HP was really tough. I loved that company and I loved my job. That “breakup” was one of the hardest for me (esp. since the other breakup happened to be with my wife, who happened to be on the other side of the civil war (in marketing) and bringing the war home at night didn’t do us any good). But it was probably also my end of innocence about business life. HP was a cult, not just a job, and when you fall out of a cult you lose a lot more than just a job title. I never imagined, that first day I showed up for work, that HP wouldn’t be my entire working life and that dreamy first year in Palo Alto and HP I thought would never end. But it did. And it’s shows how nothing in life is forever.