The three ways HP has screwed up

WHILE THE question of Alphacide is a religious debate beyond the scope of this article, HP showed inconsistency after its death. HP delivered a roadmap that showed that the Alpha EV7 line would be the last of the line, and several variants would be produced. There would be support, spare parts, and lots of tender loving care for those people who bought HP all those planes through exorbitant margins. More importantly, there were these nifty Itanium boxes that would take over when the customers were ready, with familiar OSes and support. Porting would be a snap, and customers would notice no disruption. In fact, there would be several chips between now and then so that you could upgrade for more performance.

But things started getting chipped away. New chips started getting scaled back further and further. The final ignominious blow was that the EV79 variant was summarily canned and replaced by a vastly inferior EV7x. HP didn't apologise but rubbed salt in the wound by crowing about how it kept to the roadmaps and delivered something, even if it was pretty much nothing.

The problem is that customers, as moronic as HP thinks they are, are really bright people. They understand something that HP perhaps does not, and that's business continuity. Effectively, HP was sending a message to its very best clients, and it went something like this: "We said we would do something, and on paper we will. If you are dumb enough to try and follow it, your life will be pain. Leave now. Can we sell you this nice Itanium box?".

This was not a good situation if you value business continuity, and the companies we've talked to got the message loud and clear. One major financial institution told me about the planning meetings they were having. When HP announced the Alphacide, this firm took it in its stride. It planned to migrate to the next big thing HP had to offer late in the decade. By then, the new boxes would be out, tested, debugged, and there would have been plenty of time to move its internal applications.

Then came HP retreat after HP retreat. Planning meetings went from new HP boxes in 2007 to no more HP after 2006. 2006 became 2005 if possible, and by the next meeting, HP was shown the door by the end of 2005, sooner if humanly possible. This Global 1000 company was not alone and its scorn was well deserved.

Storage
Then came storage. Read these articles if you are not familiar with the storage story so far. 1, 2 and 3. The nub is HP was canning their internal storage arrays, the EVA line. If you ask HP, it will deny it all, but the engineers have disappeared, so it is pretty hard to see how it is going to do anything new.

The roadmaps went from EVA, the current generation, to EVA-XL, the next one, to EVA-YL, the last of its line. Then came GenEVA, the next breakthrough tech. It was a good plan, and from what I am told, filled with solid if non-remarkable products. People counted on buying a refrigerator full of hard drives, and then having a bigger, better, faster one to replace it with. Rinse and repeat.

HP killed all of that with one swipe of the pen. Then they told clients that the INQ was full of it, and the next gen, EVA-XL was right on schedule. It either re-purposed the badges on EVA boxes to read EVA-XL, or it's being economical with the truth. Either way, EVA-XL is not what people were promised. EVA-YL is dead if you are an account that warrants an honest answer, but some others, it appears, are not worthy yet.

In two tries, HP went out of its way to drive a stake through the heart of its market. Once again, we received lots of letters from new Hitachi and EMC customers, and as soon as HP has the guts to admit it, or more likely, the customers wake up and realise how there are being played, there will undoubtedly be more defections.

The Tru64 Debacle
Could it be worse? Sure it can, this is the New HP we are talking about after all. Lets take a look at the OSes that tie them together, specifically Tru64. Tru64 was the old DEC Unix, bought with Compaq. It had some features, most notably clustering, that no one else could match. When you read about it, there was rarely a bad thing said. Maybe people were not too keen on the hardware or the price, but technically speaking, it was pretty rare for people to badmouth it, mainly because it worked, and worked well.

HP of course, killed it soon after the Compaq merger. Customers howled, but were soon placated by a roadmap that showed a smooth transition to HP-UX, with all the nifty features that they knew, loved, and depended on, making the move with them. It would simply be wonderful, you could not ask for a more orderly move, and customers would be coddled for the millions and millions they spent.

It was a shock to see no one paying the least bit of attention when HP canned all of the transitions a few weeks ago. Yet another promise delayed and delayed and then killed without regard to the customer. Whatever the excuse is, it won't matter, because customers have got the message.

Sun mocks
Zero for three. What a record of success for HP to stand on. It is such a sad situation that Jonathan Schwartz is mocking them in his blog. The thing that makes me know that he is right is that HP is going after him.

Here is a tip for all you HP watchers out there, when you are right, and you expose all these shenanigans , HP does not admit it, it attacks you. You should see the stuff we haven't published. Classic mismanagement, finger pointing, and head in the sand top down profiteering.

Basically, every time HP has had the opportunity to do right, it has gone out of its way to do it wrong. I honestly do not understand why HP is doing this. Incompetence is one thing, but this goes way beyond that into the realm of masochism. It can't be money, because all of the divisions that are being run into the ground were very profitable.

Now that you understand the roadmaps, why deviating from them is bad, and how HP is doing it, here is one final nail in the coffin of HP's enterprise hopes and dreams. HP support going into the toilet is nothing new, but it has recently taken on a whole new dimension of stupidity. When a firm buys a Presario and it falls apart, it has now been conditioned to accept a poorly trained outsourced person reading from a spreadsheet. Solutions to problems have become a bonus, not an expectation for a modern HP customer. If you are transcendently lucky, you may get someone who actually has a clue, or passed on to someone who has first hand knowledge of the product. If this happens to you, feel blessed, and not because it saves HP money, but because you are truly lucky.

It's myopic mismanagement that puts saving pennies ahead of long term profitability. While we all know cost is not the main reason people buy multi-million dollar boxes, HP still does not appear to understand this, and has outsourced the support for several key enterprise products.

Dell, for all the vilification heaped on it, recently started to take its high end support back onshore. Not surprisingly, customer satisfaction rates jumped as soon as it did this. Taking care of your customers is good for business. When you spend millions of dollars on support response with times measured in single digit hours, you expect to actually get service.

The number of letters I've received telling me that they called HP and got someone on the line who does not speak their language natively and has never seen the machine they are 'supporting' is simply amazing. It may save HP tens of dollars here and there, but surely it is costing them millions. When you are losing millions of dollars an hour, repeating things over a long distance line is not comforting. When that person gives you the Superdome equivalent of 'is your unit plugged in sir', you tend to go ballistic.

The kicker is that if you actually convince them that there really is a problem, there is no one to pass the call to. They fired them all. Really. Go to the buildings where the Tru64 staff once was and look at the cars in the parking lot. Do the same for the storage division. There is no one left to fix the problems that did arise, much less the ones that will.

HP does not seem to grasp the concept that you cannot do a credible, timely job fixing a problem with a business critical system if you have never looked at the code before. The person who wrote a subroutine can probably hunt down and fix a bug in a timely fashion, a person handed a phone book sized printout that they were given three weeks of training on just doesn't have a chance.

The 10K foot view is simple, it has simply betrayed the trust of its most lucrative customers. It did it with products and with support. Customers are running screaming, but due to the nature of the products, they can't leave quickly. But make no mistake, they are leaving.

Any sane evaluation by a potentially new customer would be equally problematic for the company. All the promises of future roadmaps, both hardware and software, would have to be evaluated against the track record. Not a good start to a relationship.

Then comes the products. HP has gone from designing its own CPUs, its own OSes, and their own storage arrays to being a reseller. Why would I want HP, a company prone to snap changes on long term roadmaps, to be a part of my long term strategy? In fact, if I wanted cluster technology, why would I not go to Veritas? All HP does is resell them, so why not drink from the tap. It also conveniently avoids the whole support with no product knowledge trap.

For OSes, you go with Novell or RedHat, MS doesn't tend to play well in this arena. For the chips, there is no shortage of people willing to sell you Itanium boxes. You can get everything HP sells from many other vendors, and this is ironically enough the New HP strategy, resell.

The reason you would go with HP is simply that it is the best. It inspires trust and confidence, you know it will be there for you. It is worth the extra money for the peace of mind that your business will be in good hands when you need it.

If I have not made it perfectly clear yet, HP has spent much of Carly's reign destroying this trust. The most problematic part is that it did it the right way, with action, not words. If it had stated it was not going to live up to its promises, it could always fire token people, recant, and beg forgiveness. When you fire the only people capable of doing the work, and the customers get poor service, sorry does not cut it any more.

The well of high end, non-x86 based devices is poisoned at HP. In my opinion, the division is pretty much a write-off. You cannot reconstitute a team with 20 years experience developing your products. There were people who did it, and there are others. Money will not buy the necessary talent, and it is pretty clear that the bridges to the needed people are burned. There is nothing left. Billions have flushed needlessly, it may take a bit for it to be noticeable though. Welcome to the new HP, this is a 'victory'. Bonuses all around. µ