What Were They Thinking?

I guess I don’t have to worry about giving away any secrets regarding the strange behavior of HP in the last few months, because developers like me haven’t been told anything that isn’t available publicly. Maybe I’m a little more compulsive than most in collecting that info, so I thought it might be interesting to assemble the story so far as I understand it.

1. The board has decided to get out of volume/consumer businesses, or at least deprioritize them, in favor of complex-sale/enterprise markets. This was decided perhaps as early as the hiring of CEO Leo Apotheker late last year, formerly of SAP. (i.)

– This is sometimes miscommunicated (by Leo and the board) as concentrating on software rather than hardware.

– While there are entrenched players in structured data (i.e., pseudo-relational databases), and maybe even market saturation, the board feels there are open growth opportunities in un-structured data.

2. HP has three divisions:

Business Critical Systems (enterprise, including servers),
Imaging & Print Group (printer cartridges made by others, but also scanning), and
Personal Systems Group (PCs, and also Palm until things got weird).

– BCS is already by far the biggest enterprise vender. (ii. iii.) HP will use that, um, leadership, to develop the unstructured data market, and use IPG to scan all that unstructured data.

– PSG is the largest PC maker by volume, by revenue, and by profit. While it doesn’t fit into the above, it is supposed to be able to stand on its own. So it will be split off, just as the old test/instruments group had been split off from HP years ago as Agilent. From the Agilent experience, the board knows that this will take 12-18 months to accomplish, and that there is even a fiscal quarter in which they’ll have to exercise due diligence on alternatives such as keeping it or selling it. The former is just pro forma (as the board knows it doesn’t want to keep it), and the latter is unlikely (as no one could afford to buy it).

3. The above plan is bolstered by buying the largest software company in the UK, Autonomy, which specializes in unstructured data for enterprise. (iv.)

4. HP is infamous for leaking plans by accident. It happened with last quarter’s financials, which is a big SEC no-no as it allows people in the know to make money that the rest of us cannot. It happened years ago with the board ending up illegally investigating each other to try to find the leak. It happened again this time. No one was supposed to know any of this — including the business unit executive vice presidents — until the plans were finalized. Once it was leaked, the board had no choice except to announce everything, even though the plans were not complete. HP had already lost nearly half its market value since Leo had taken the helm less than a year before. After this debacle, it lost another 20% overnight. Now it is so cheap that it might be a takeover target by Oracle or some such.

It isn’t clear to me at all how Palm fits into any of this. Leo’s previous plan, just a few months old, was that BCS would sell webOS devices throughout the enterprise, including tablets, phones, printers, and PCs free of MS Windows. Now it is not clear if webOS will be spun out with PSG, or retained by HP, or sold off to someone else.

Although there is a lot of industry consensus that webOS would make a fine alternative to iOS and Android, it is not clear to me how it would actually fit into anyone’s business plan. E.g., not hardware makers HTC or Samsung.

I do feel that there are a small set of companies who would love to acquire 500 really good mobile engineers all at once, together with an excellent mobile patent portfolio. But that would have to happen fast before we all leave, and I imagine that ship is leaving the dock now. Or it could just be sold for the patents. Palm is perhaps the one tech company that Apple has never sued, and has one particular patent that is reputed to be THE patent for smartphones. (http://www.google.com/patents?id=pJnHAAAAEBAJ&printsec=abstract&zoom=4&source=gbs_overview_r&cad=0#v=onepage&q&f=false)

Google and Microsoft make sense in the long run, as they both need a mobile platform based on Web technology (e.g., not based on Oracle’s Java nor as isolated as Windows now is), but for both it would be so disruptive in the short term that it would be better to go with any of the alternatives already being developed by each. Nonetheless, I imagine that Google is bidding something silly just to be “at the table” and disrupt the process, as they did with the wireless spectrum auction and the Nortel patent portfolio.

There have been articles about how Amazon would benefit from rolling their own on a more advanced (if not more mature) base than either Android or iOS. There have been articles about how Facebook could change the entire industry by having their own app platform that works on both devices and desktop browsers. Either works, but any such sale should have happened by now. I imagine HP screwed up by trying to sell the one OS to both instead of either, while simultaneously keeping the use of webOS for themselves. After all, being an end-user computing device maker, PSG will certainly need a go-forward plan for a Post-PC world. (Windows 8? Really?) Plus Amazon and Facebook do not compete with each other. But could  you imagine a consortium with three major players? It would be like playing Monopoly with three people: the weakest of the trois invariably hooks up with one of the others against the third, and we all know who that would be. In any case, who on earth would want to join a consortium with HP under any arrangement today?

I don’t know what Leo and the board were actually thinking before, or what they are thinking now. One school of thought is that Palm was basically just too small to matter. New CEO Meg Whitman has expressed her horror at learning what Leo did to Palm, and yet she was on the fucking board while it happened. John remembers telling a CEO years ago that the guy was “gambling with his own beer money, but with [John’s] rent money.” They guy laughed at John in response. Or maybe Leo just folded his cards when Google went “all in” with his chips by shockingly buying Motorola less than week before HP rushed out its ill-prepared plan announcements. Anyway, I don’t buy either of these. It doesn’t really make sense to kill off a $300-in-parts tablet that actually sells out in a weekend at $300, while PSG keeps making $1000 slates that don’t sell at all. Maybe by killing the device but not the OS, Leo was trying to make a point about software vs hardware, and just got the messaging all bollixed up. Or maybe it was just an expensive “Fuck You” to former CEO Mark Hurd, who had bought Palm for HP in the first place.

In any case, there was certainly something weird going on long before the announcement that HP would stop making Palm hardware: for the TouchPad launch, a complete idiot of a sales exec was transferred out from HP into Palm, and he really screwed up — sales teams were not trained at Best Buy etc., display’s not ready and filthy dirty, pricing chaos, retailers pissed at the chaos and lack of support from HP, etc. They changed the name “Palm” to “webOS” and redirected the traffic, but didn’t bother to obtain the rights to the domain “webos.com”. At one point before that Thursday when everything went to hell, there were three different sets of TouchPad prices listed on HP’s own Website. Third party developers were not given tablets. Software partners were not ready, including the one providing office-doc editing or the one providing streaming TV shows and movies. All webOS apps were supposed to run on PC’s, but the software (from HP, not Palm) is still nowhere in sight. A Christmas deal with Walmart failed to materialize. A cute 7″ tablet with identical resolution was leaked and ready to produce, but was put on hold. Phones were announced and ready but put on hold. Totally Osborned. Maybe the whole Palm purchase and withdrawal was an elaborate “New Coke” move to negotiate better terms for Windows 8m which HP now says will be used in PCs and Tablets going forward.

I expect we’ll see Meg and the board blame everything on Leo, and try to walk some of it all back somehow. Watch for someone to make cracks about the Apotheker being on drugs. But I feel it’s too late to walk Palm back. I feel pretty crappy about it all. We were told HP was totally committed. We were told this was a marathon not a sprint. We were told HP would be hands-off for another year yet, except for marketing. We were told about <a href=”http://in.reuters.com/article/2011/10/21/idINIndia-60032020111021″>Invent</a> and <a href=”http://blogs.hbr.org/taylor/2011/09/how_hewlett-packard_lost_the_h.html”>The HP Way</a>, and that everything at HP depended on leveraging suppliers by producing every kind of computing device at every volume and margin scale.(v.) But mostly I’m pissed that I had decided that enterprise software was no longer the driver for technology innovation, and that consumer mobile is, and that the CEO of the world’s biggest tech company has now told me that this is a really dumb idea.

——–
i. Leo had been fired from SAP less than a year earlier for raising prices during a recession and nearly ruining the company.

ii. Of course, a lot of HP’s high margin results are actually from banking services, in the same manner as GM being the largest car maker by virtue of GMAC, or GE being being so big because of GE Capital. So volume and profits in this area are really inflated.

iii. BCS’s most pronounced leadership is in owning the market for a server architecture called Itanium, which Intel makes the chips for, and for which Oracle makes most of the software for it. Microsoft and other venders dropped support a while back, and so HP asked Oracle to declare it’s love. According to Larry Ellison, that made him suspicious enough to ask Intel what the story is, and learned that Intel wasn’t going to continue major development on it. (At the same time, Larry had just hired ousted HP CEO Mark Hurd. So “maybe” he had other sources of info. In any case…) Larry claims to have been shocked, just shocked, that HP knew Intel was dropping Itanium and didn’t tell him or his customers, and so Larry responded by dropping future development for HP’s lead architecture. (Itanium competes with SPARC, that Oracle now owns, so there “could” be other reasons for Larry to drop the HP server.) Anyway, sales of Itanium has dropped like a stone, and it is difficult to see the way forward for BCS servers without some serious long-term investment.

iv. HP is paying an incredibly high premium to buy Autonomy: more than 10x revenue and more than 45x earnings. Apparently Leo and the Autonomy founder go back quite a ways. Not clear yet if Autonomy has some sort of patent protection for their otherwise unspecial business, but I do imagine a software patent war coming. Meanwhile, to raise the cash, HP will have drop it’s barely started plan to boost it’s stock price by buying up shares. It is not clear yet what other conditions bankers may have imposed for providing the cash (e.g., getting rid of business that aren’t producing a high enough return next quarter). Palm doesn’t have any financial services to boost its apparent profits. There’s also some sort of silliness going on again with Oracle, as Larry claims to have turned down a seduction by Autonomy, which HP/Autonomy denies, and which Larry then has claims to have the proof for on paper. I don’t know what that’s about. Also, at some point Autonomy produced a consumer mobile app called Aurasma. Really. I’ve tried it, and well, they’d better stick to enterprise.

v. HP is known for its “supply chain management.” What this really means is squeezing suppliers and assemblers. The idea is that when you make lots of different stuff, you can get a good price on volume parts by offering a vendor a contract on some other high-margin part for a different device that you also want to make. Supposedly, this applied throughout enterprise servers, printers, PCs, phones, tablets, and accessories. It is not clear how any of that will matter in the new HP, although one exec pointed out that there are “plenty of industries that have such economies of scale among companies of different ownership.” I don’t know. Maybe that means a Keiretsu in which the HP board can tell an “independent” PSG board what to do. In any case, the month before the change, execs extolled the virtues of HP supply scale. After the change, the same individuals tell us that “for several years now, HP strategy has been based not on trust, but on contracts.” Whatever.