HP shrank its base, so you grew experience
November 17, 2010
Second of two parts
We don't let a November pass without insight into the HP 3000 exit announcement which HP GM Winston Prather pulled out like a motley magic rabbit in 2001. Prather and HP have stuck to their story over why it's ceased development and sales. But while the exit of much of the 3000 business at HP is only weeks away, it's taken years to understand another story, told by more than HP and then parroted by the Interex publication, HP World.
2001 was a time when the user group was courting HP's alliance in radical ways, like branding its conference as well as its monthly news publication with a name that HP owned. Ever the lapdogs of casual journalists, analysts were called in at HP World to confirm the pullout was good business. For somebody, anyway. A Gartner analyst said in the HP World opening story about the migration, "If HP moved MPE to Itanium, they'd probably lose most of the [software vendors] anyway. It's a fine line between taking care of customers on the one hand, and making money and staying in business on the other."
If you can hold off on the horselaughs for a moment, try to understand what the Editorial Director Mike Elgan was printing, without any opposing viewpoint. HP needed to stop taking care of these customers, or it was going to go out of business. Even back in 2001, you couldn't find the total sum of HP's 3000 R&D plus sales of the 3000 without a microscope. The company was doing $45 billion in sales that year overall. If the 3000 group represented even 1 percent of that total, we could never find the company saying it. Those kinds of product line numbers were HP's secret back then.
Those partners who were deciding the 3000 wasn't strategic? One, the exclusive North American distributor for the 3000, reported in the NewsWire that it broke all of its previous monthly unit sales records during October -- doing all that during a weak IT economy. "Dan Cossey, Senior Integration Engineer for Client Systems, said the pace of rolling out new systems has been good for reducing waistlines, “We’ve been so busy building systems that we’ve got to force ourselves to take a break just to eat,“ he said.
That's the same Client Systems that HP World predicted "will likely be busy over the next five years with migration work." Another HP voice assured us that Client Systems was "stepping up to take a large role in the mass migration." As it turned out, the years 2002 through 2006 had large roles for software companies and HP Services, but not much software stripe moving at Client Systems. HP's experts presided over a $14 million meltdown of one migration that resulted in a multimillion settlement and a restart by software experts.
Like the 3000 vets fingering the architects of this mess with a clever mascot and shirt motto, who ever knew about the 3000's future? The customers, the engineers in HP who had to follow orders or leave jobs. If HP had moved the 3000 to Itanium... if the new hardware models, just six months old when Prather swung the axe, got a chance to gain converts... if the hacking away HP products wasn't in full flay in the months before that November nadir of news... things might have turned out differently. HP might still be migrating customers away from the 3000, but it would have uncorked an announcement with full-staffed programs, tools built for system and app experts — a way to take care of customers who'd had no intention of moving away from a server built to keep a smaller firm from building a massive, churning IT datacenter. They'd grumble and have to learn quick, but the vendor could help retrain them, then retain them with goodwill. Today, more than half who've left won't even boot an HP operating environment at gunpoint.
In short, it would have been a novel story to tell about ending a business that supported tens of thousands of companies and turned HP into a computer company. Those thousands, that was a number provided by analyst house IDG one year after Prather's decision. Instead of dreaming up policy and education and engineering, playing catch-up as it went, HP could have been stewards of responsible change for anybody who could afford to put change into practice. Finding a big chunk of surveyed customers "planning, or partially implemented, strategies for migrating" in the HP World story meant very little to anyone -- except a system vendor who needed to end a business line not as large as some of its others. Including one owned by the company that HP wanted to acquire. Whoever really signed off on the 3000's HP fate, no Compaq exec was ever nudged to put OpenVMS near the chopping block.
Prather and the HP upstairs boys were guessing that the company's failure to sell 3000s was going to continue and accelerate. He had to make a decision, his decision, so HP could bail out looking like that was its forecast all along. It just didn't sound like the plan he offered just two months earlier in Chicago, at the user group's convention.
He asserted the A-Class e3000s were more of a success than even HP projected. “We project what we’re going to sell, and the A-Class systems are selling at a much higher rate than we anticipated,” he said at the HP World show. “We’ve seen the new systems taking off much faster than we projected, a more rapid switchover.”
Six weeks later, somehow, the system had stopped being strategic to the partners that Prather brought onstage to thank for being so special, as well as the customers he reported were buying up the servers rapidly. The slide below from August 23 shows how HP spoke in 2001 and yet didn't address the real future, the system's software. These "platform" plans were all about hardware, already in the pipe.
So in this week you might as well go ahead and admit it. You know it was your fault that Prather, and the HP brass jamming a merger down shareholders' throats already rattling with resistance, had to make this decision. A "difficult decision not only from a business perspective," he said, "but also for me personally." Well, maybe not so difficult personally by now.
As we noted awhile back, Prather — who retired the 3000 GM post when he left the group -- enjoys a GM position leading the NonStop computing products from HP. Lab people inside the 3000 division said his exchange with the HP leadership was to get a transfer or some HP work for everybody in the 3000 group, in exchange for swinging a sword he claimed to own. But of course the HR blood-leeching kicked in pretty swiftly at HP before much more than hardware deals and Webcasts could be produced. What was promised could never be delivered to protect many MPE tech pros inside a vendor gone crazy for PCs and printers. Retired, some; released, others. A few still reach out to help you, after hours.
This month we are thankful for the brilliance of independent 3000 pros who are mucking out those stalls which that 2001 decision filled up. The tools are on hand, the experience is finally available, and hardly a lick of it comes from HP. At the Washington State Board for Community and Technical Colleges, its migration project leader was plain about it. "HP hasn't given us a lot of support here," he said. "We had to locally hire our own HP expertise." We couldn't find any adoption of HP-UX applications for the City of Sparks, either.
Despite that rugged road, this week you've moved into the 10th year after that Prather decision, or about twice as many as he predicted HP would take to get most of the community onto its other computers. Things have turned out differently over 10 years. Back when this mistake was made, Palm Computing was king of what we called PDAs. You might know them now as smartphones, but few were smart enough to swap data across the air. The D stood for Desktop.
Their reach into a booming Internet world didn't matter; these were going to be important tools of the future, according to Editor Elgan. "The mighty Palm organization has been dethroned by the HP Jornada," he wrote in a column just two pages after announding the Exodus. "I still believe Palm organizers are for novices and cheapskates." This, after he reported he'd once published a newsletter called "The Palm Reader." Like Elgan and his cheapskates, you might have heard of this Palm place, the one HP now owns and whose former CEO leads HP's PC group. Again, it's the software, Palm's WebOS, that shapes the future.
Elgan's Palm obituary was only about 30 percent shorter than the parroting about the 3000's exodus. People still believe the 3000 was a doomed product during that fall nine years ago. But it was more like damned by HP than doomed, milked out for its enterprise strategy innovations and then branded as bereft of a future as today's Jornadas. Or any further insight out of Interex, a group whose PR director advised me in 2003 we'd better get a fresh business model to survive, if covering the 3000 was our only business. She wasn't available for comment in our story about Interex's overnight bankruptcy two years later.
I'm not even planning, let alone starting to implement, any strategy to act on her business advice. There's too much good work to be done "Helping IT Professionals Manage Change," as HP World proclaimed on its banner. Someone has to get the work done to make changes, instead of waiting for HP tell even more Amazing Stories of the future. There are some things that have died out here in the future, yes, but your ability to make a 3000 support a business isn't one of them -- unless you tell the tale with your own facts from the field, instead of a corporation viewpoint.
Happy Anniversary, 3000 community. Perhaps we'll all gather somewhere next year to raise glasses as we did in 2003, and so celebrate the Eighth Anniversary of the system's World Wide Wake. That will be in the 39th year of MPE service, and the first year of its survival after the end of HP's 3000 life. Cue the scary organ chords. It's the afterlife, after all, where IT professionals managing change are the most genuine source of stories. From what we hear out of them, they're the ones making the decisions, working out the number of their 3000's days for its future.