Pattern Finder

HP's Hurd

HP announced today that Mark Hurd of NCR will be taking over Carly Fiorina's job.  The New York Times take is that it's a good fit; I agree.  I've met Mark and saw him operate at the Teradata division of NCR.  The two companies are culturally quite similar; quality engineering is central to both.  They both have the attitude that superb engineering will win out; marketing, while not ignored, is somewhat secondary.  HP under Carly tried to give more gloss to the marketing, but I think inside the company she was despised for it.

Mark's biggest challenge will probably be getting used to the Silicon Valley ambience.  Trust me, Dayton, Ohio (NCR's headquarters) is not known for its fruit smoothies or veggie wraps.  So hurrah to HP and condolences to NCR, which I know will miss Mark terribly.

Google Buys Urchin

On Monday, March 28th, Google announced that it had purchased Urchin Software, a Web analytics ASP.  The terms of the deal were not disclosed.

Urchin (previously known as Quantified Systems) was one of those quiet Web analytics companies, never one to toot its horn much, although it was a founding member of the Web Analytics Association.  Run by a bunch of folks who'd all graduated from the University of California, the company always emphasized its emphasis on scalability.  Besides selling to enterprises, the company sold to ISPs.  Culturally, Urchin is probably a good fit for Google, as I was always struck during briefings at how engineering-centric Urchin was.  This is usually a long-term minus for a company, since if you don't marketing, people don't know how good a mousetrap you have.  However, since Urchin is now part of Google, the visibility problem instantly goes away.

Well, one more datapoint for my prediction that Web analytics is becoming a part of the online suite (content management, search, personalization).  WebSideStory purchased Atomz recently, and Plumtree just announced that they were offering embedded Web analytics with their portal software.  As an analyst, given the hits that the IT analyst industry sometimes takes for blowing hot air, it's nice when the future unfolds as you predicted.

The Shoe Finally Dropped

Well, the event that the Web analytics market has been waiting for finally occurred:  NetIQ has spun off WebTrends, which is being taken private.  According to the NetIQ press release, "Francisco Partners will acquire NetIQ's WebTrends business unit for approximately $94 million in cash."

WebTrends had been a self-sufficient division of NetIQ (with its own set of financials) for awhile, so the analyst community had been surmising that WebTrends would ultimately return to being separate (as it was before NetIQ bought it).

By going private, WebTrends get to avoid the overhead of Sarbanes-Oxley, as well as the ability to call its own shots and acquisitions without Wall Street looking over its shoulder.  Given that WebTrends management believes that the time of consolidation is here (a la WebSideStory's recent purchase of Atomz) it will be interesting to watch things unfold over the next year.

Stellent, Are You There?

The Stellent Web site (www.stellent.com) was down this weekend and is still down as of this morning.  Should we read something sinister into this, or did someone just forget to reboot the server or not pay their DNS bill?

When I was at Aberdeen Group, we vanished off the Web one day.  We didn't cotton on instantly, but by the afternoon of that Monday we started getting calls from clients asking, "Why isn't your Web site up?"  Since the server was humming along just fine, we couldn't figure out what they were talking about.  And when you connected to our Web server via its IP address, it was clearly up.

However, our Domain Name Service records had vanished.  So when someone typed www.aberdeen.com, they were never forwarded to our Web server.  It turns out that our DNS provider had suddenly decided that we hadn't paid our bill, even though we had the cancelled check (that they had cashed) and the resulting deposit record.  Even faxing those documents to them didn't help.  However, as an analyst firm, we had a trick up our sleeve.  We started calling newspapers with the story, and when the company got bombarded with calls from reporters, it suddenly decided to be a lot more customer-centric.  Even though they fixed the problem the following day, it still took the good part of that day for the revised DNS records to cascade through the system.  So Aberdeen Group was offline for most of two days due to a third-party clerical screw-up.

Just one more thing to worry about in this online world.

Endeca's Latitude

Well, Endeca finally got around to announcing Endeca Latitude.  The software has been around since December, but the company apparently figured they'd get a bigger splash if they announced it at Esther Dyson's PC Forum.   (Having met Esther a number of years ago, I understand the thinking.  And boy, does she enjoy fresh cherry tomatoes.  But that's another post.)

I wrote about Endeca Latitude in a January Briefing Note (warning:  the full text is only accessible to Ballardvale Research clients, but you can read the summary here).  Latitude is a query front end to both unstructured and unstructured data.  Its beauty is that workers that can't be bothered to learn a business intelligence application (e.g., Business Objects, Cognos) can probably find the information they're looking for via the Latitude portal.

As I note in the Briefing Note conclusions,

Endeca’s Latitude is one example of the quiet revolution occurring in online information retrieval. In the past, BI applications and portals were the engines of choice, largely because structured data was accessible online and unstructured data (memos stored in file cabinets) wasn’t. However, almost all corporate data is now created digitally -- via Microsoft Word or Excel, or in e-mail -- so structured and unstructured data are now equally available. Going forward, the information retrieval winner will be the application that can mine both data types equally well -- and so far, enterprise search companies such as Endeca have done a more even-handed job than their BI counterparts. Companies should not necessarily de-install their BI applications -- but they should investigate search as a way to make BI and other corporate information much more accessible to large numbers of employees.

My more in-depth musings on the subject reside in my December 2004 DMReview.com column.  No matter how you slice it, information retrieval is going to get easier in the coming year.

 

I Can Tell It's Spring

Well, it's Spring, which means Web sites blossom with press releases about how "Company X has been named to KMWorld's 100 Companies That Matter List...."  Examples include Attensity, Autonomy, EasyAsk, Ektron, Entopia, Interwoven, Inxight Software, ISYS Search Software, Mondosoft, Percussion Software, and Red Dot Solutions.    

OK, self-congratulation is to be expected -- and its nice to have something to talk about when Development hasn't shipped anything in 6 months --   but what does it really mean?  Are there really thousands of Knowledge Management companies?  One hundred is pretty close to the number of viable KM companies out there.  The Top 25 would be interesting; All the 100 is not.  Realistically, the award should be renamed to, "KMWorld's 100 Companies That Are Around."

Business Objects XI

I had the opportunity to see Business Objects XI in action (as opposed to screen captures) recently at a rollout seminar and was quite impressed.  Business Objects, like a number of its competitors, has finally realized that Microsoft owns the desktop, rather than the BI vendors.  Consequently, Business Objects has worked at integrating its software into Microsoft Office.

There are, of course, varying levels of integration.  Everyone exports to Excel these days; you have to be stupid not to.  However, Business Objects does things such as using BO as a data server to PowerPoint.  The numbers aren't up-to-date?  Just do a refresh, and the numbers correct themselves.  You don't need to exit PowerPoint, start up BO, do a query, and then copy and paste the results into PowerPoint.  You can even do some querying from within PowerPoint.  So corporate users that want the insight of BI can have it -- but without the labor of learning another application and user interface.

In fact, it would have been a 5-star out of 5-star morning, if the VP of North American Sales hadn't opened his mouth.  Driving home the fact that he was the go-to guy, he queried the audience on the presidential quote regarding accountability.  "The buck stops here," was the answer from the audience.  "Absolutely," said the VP, referring to a placard which Harry Truman kept on his desk, "That quote from FDR is famous."  So the folks at Business Objects clearly know technology, but they don't know their history.

Eliyon Technologies

Some business plans you just wish you'd thought of yourself.  Take Eliyon Technologies for example.  This company scrapes the free information about people off of Web sites -- from press releases and the Management section -- puts it into a database, and then resells it.  Over time, it's created quite a large database that headhunters and sales reps salivate over.  Trying to hire a Marketing VP who previously worked for SAP?  No problem.  Trying to figure out where the decisionmaker worked before?  No problem.

However, the part of the product that takes the cake is that you have to pay to correct their mistakes.  For example, Eliyon had me at Aberdeen Group (where I haven't been for a year) and listed me as a Senior Analyst (even though I'd been promoted to Research Director in 2001).  Yes, I'd been at Ballardvale Research, but before I worked at Aberdeen.  Needless to say, I wanted to correct this.

So I logged onto their site, spent about five minutes fixing my titles, order of employment, etc., clicked OK, and then was notified that since I was a "public figure" (they'd found me more than 30 times on the Web) I had to prove who I was to prevent fraud.  So I had to enter credit card information and get charged a buck (solely for the cost of the transaction they claim) before they'd let me correct their mistakes.  (I find it hard to believe their costs are exactly a dollar.  Maybe 35 cents perhaps, or 75?  One dollar just seems too round a number.  Of course, if you add some profit in there, maybe it does turn out to be a dollar.)

See what I mean?  Don't you wish you'd thought of that business plan?

Low-tech RFID for the Parking Lot

The Boston Globe had an interesting article recently with the long-winded title, "Forget where you parked your car? Logan Airport is spending tens of millions of dollars to track available spaces and help travelers navigate its busy garages."   It says that cameras at Boston's Logan Airport use optical character recognition (OCR) to log license plate numbers as cars enter the airport's parking garages.  That night, employees then use handhelds to update the license information with the car's lot location.

The benefits are several.  First, parking garage management gets an up-to-date tally of how many spaces are remaining per garage level.  Second, garage attendents can tell people where their car is if they forget where their car is.  (Apparently around 20 people a day fall into this category.)  Third, Massport officials can figure out the appropriate parking charge, even if the customer loses his or her receipt.  Finally, the system deters theft by matching the license plate with the parking ticket; if the two don't agree, the police interrogate the driver.

It is worthwhile for Massport to go to these lengths because it made $74 million from parking charges last year.  But it's a preview of the benefits that RFID will offer as companies grapple with balancing inventory with demand in a realtime fashion.