HP reverts to type

Last week Mike Lynch, the founder of Autonomy, resigned from HP. We should have seen it coming. Not for nothing is HP often considered to be “the company where software goes to die”, primarily because the company appears to see software as a way to sell more tin. HP did make an effort to change its ways by appointing first Mark Hurd and then Leo Apotheker as CEO but neither managed to make the company transition to anything other than a hardware vendor.

Mark Hurd introduced NeoView but that failed. The story I heard was that the warehousing specialists that Hurd brought in from Teradata were at loggerheads with the ex-Tandem people about the architecture of the product and, unfortunately, the latter won. That meant that, for example, you had a switched fabric for connecting to disk rather than direct attached storage and while a switched fabric is a great for high availability it sucks for performance. It also didn’t help that HP didn’t understand the market: one executive at the company asked me if NeoView should be within the hardware division! I rapidly disabused him of that notion but it’s indicative of the way the company thinks.

Over the years HP has bought and effectively buried any number of companies. Where now is SeeBeyond? Or Outerbay? The only area where it has been really successful is with network management software but that’s hardly surprising given its close correlation to the network hardware that HP is purveying.

The question of course is what to expect of HP’s major software products now? In particular: ArcSight, Vertica and Autonomy. The reason for Mike Lynch’s departure depends on who you talk to but one thing is clear, which is that sales of Autonomy have fallen off a cliff as the new CEO, Meg Whitman, eschews the software focus of her immediate predecessors. Of course the danger is that a vicious circle gets established: people get worried that products like Vertica and ArcSight don’t have the future that they might be expected to have (and I admit that I am not helping the HP cause here) and therefore they choose alternative products, so sales drop, and that induces HP to reduce R&D in those products, which means that sales drop further, and then eventually the product dies.

All of this is not helped by the fact that HP has also announced that it is cutting staff numbers by 27,000. Unfortunately, as we are all too aware over the last few years, companies often make repeated announcements about job cuts, so we can’t guarantee that this isn’t just the tip of the iceberg. Meg Whitman has said that the company is going to be re-focusing on R&D but one suspects that the emphasis here will be on hardware and not software so I am not sanguine about the future of Vertica, Autonomy or Arcsight and I would think very carefully about introducing any of these products if there is a suitable alternative.