Autonomy marketing, meet HP

Leslie Owens, the enterprise search analyst over at Forrester Research, has written one heck of an article about a potential Autonomy in the HP era. Her analysis strikes me as being very insightful and, in my opinion, quite accurate. What makes it unique is you just don't see alot of 'analysts' tell it like it really is. Kudos to Ms. Owens!

Technical issues aside, I'm reminded of a story that goes back to the early days of the PC when HP and Apple were just beginning to compete. A popular quip about the difference between HP and Apple went "Where Apple sells sushi, HP sells cold raw dead fish". The implication, of course, being that HP just wasn't good at marketing.

I'll always think of Autonomy as a search technology company. Our first exposure to Autonomy was in 1997  with an early version of the DRE, the predecessor to IDOL. Back then, using vi or emacs to configure a search engine was pretty common; and no one really had grasped the importance of the business side of running enterprise search.

In search, IDOL returns pretty darned good results out of the box, no tuning required. But if you want to tune it, if you have alot of custom work to do, IDOL gets really tough to set up and configure... and it's still done using text editors to create and edit text configuration files. This may be one reason why IDOL projects take so long to complete and require such big teams of consultants. HP probably won't be changing this... they want to grow their consulting revenue!

But now, in the second decade of the not-so-new century, customers expect to use a GUI to configure, manage, and customize enterprise search; and just about all of IDOL is still 'command line based'. I think this is just one of the data points supporting Leslie's remark that IDOL 5 has not had a major update in over 5 years. Sure, they've added dozens of new capabilities... API calls, and the like... but the platform is still a solid 1990s kind of experience. "Powered by vi" was funny in 1998; not so much now.

Nonetheless, Autonomy has been quite effective because their technology is pretty darned good at finding content; and because their sales force has been aggressive in selling the product. HP will love the consulting; but will they be able to move product as successfully as Autonomy had?

What do you think?

 

 

Comments

[Andrew]

My experience with Autonomy and the Autonomy Business Console have been less than stellar. The relevancy of the results is so bad that I could spend the rest of time creating hypernyms and synonyms to tune them-and I probably will since the Autonomy Business Console is so slow, kludgy and crashes constantly. On top of that, support from HP is abysmal. I don't know how much of the problem is Autonomy vs. our configuration of Autonomy, but it doesn't really matter at this point. If I had the option of switching to anything, particularly a SaaS model, I would do it in a heartbeat.
==
Thanks for the feedback, Andrew! Yes, the console was a great idea.. perhaps poorly implemented. 'When will they ever learn?..' applies to more than just Autonomy!

[Ken Stoltz]

I'm not sure that's the impetus for Leslie Owens' comment about the major updates. If it's about meat and potatoes functionality, then I disagree with her. They've provided substantial changes throughout IDOL 7. Regarding the lack of Administration UIs, Autonomy has Autonomy Business Console and Autonomy Collaborative Classifier to handle query and relevancy tuning, and it's used on numerous customer sites. I'd argue that the learning curve of many Administrative UIs are steep (such as those for FAST) or incomplete as well.
==
Thanks for the comment Ken! I'd agree to get ALL features almost any enterprise app console is a steep learning curve. But being able to point type a URL, click a few buttons, and get a collection that is searchable should be trivial. In my experience it is on most of the successful enterprise platforms out there. That said, once you get it configured and fire up the right URLs, IDOL gives you pretty darned good results without any tuning... THAT's one of the things I think they do best. And in the real world, that counts alot! /s/Miles

[Dave Haucke]

Miles, your last point about HP loving the consulting raises an issue that no one seems to be discussing -- namely, customer service. Yes, HP will love the consulting, but how will they get Autonomy to suddenly be customer-driven. Each OEM partner we sign away from Oracle or Autonomy all say the same thing ... impossible to work with. That's good for ISYS, but not good for HP's customer base.

Hopefully, the deal will translate into the kind of superior customer service that customers are demanding from IT vendors today (and which we pride ourselves on), but it could take years for cultural integration of the two firms to take hold.
===
Dave, yes.. great points! As there were different kinds of company culture than MS and FAST had. This will be culture shock - hopefully it will work out without too much pain for customers.. and the sales teams!
/s/Miles