Pattern Finder

Often Forgotten -- Online Visitors Are Actually People

Eric Peterson had a post on his Jupiter blog this evening which took issue with a petition proposing "Internet Advertiser Wakeup Day," in which folks are urged to generate bogus registrations at media sites such as The New York Times, LA Times, etc. on November 13th.

I kind of got a chuckle out of it; Eric was more affronted, pointing out that gathering minor demographic data is a small price to pay for content.  The reasons for my chuckle are several.  First, if I were The New York Times, for example, I would throw out the registrations on November 13th and get on with life.  No great harm done.  Second, I thought it was a pretty creative way to point out the problem with much of online behavior monitoring these days -- there's no clear tradeoff of value for the user.

When I first visit a number of newspaper sites, I have to put in my age and ZIPCode before I'm allowed to see the content -- for no clear additional benefit on my part.  It isn't like I get stories tailored to my location or age group.  This is exactly why people put in bad data -- and the petitioner is urging them to do so -- it adds a bit of a zing to an otherwise dull day.

Online enterprises have gotten so enthralled with visitor stats that they've forgotten people are behind them.  People who are bored, busy, confused, and who aren't looking to buy the thingamabob the Web site is trying to cross-sell.  If you strolled up to a person, asked them a bunch of questions, and refused to answer any questions in return, you'd be labelled a jerk or worse.  Yet Web sites do that every day.  Until Web sites concentrate on offering better service in return for customer information (Amazon.com is the classic example), they shouldn't be surprised if the humans at the keyboard don't "play by the rules."

ClickTracks' BidHero -- and Less Is More

I just put up my Briefing Notes (clients only; summary that everyone can read is here) for ClickTracks' new bid management software.  As I note in the Conclusions,   

Once again, ClickTracks displays the design philosophy that has distinguished it since its founding: creating a tool that makes it easy to manage the essentials of a complex problem. This is not as simple as it sounds: it means avoiding over-simplification, as well as deciding what’s crucial and what’s extraneous. It’s akin to being a superb editor -- an anomaly in the software world, which is inclined to think that running on and on with features is a good thing.

While this has been John Marshall's (the CEO's) mantra since day one, other software companies are finally starting to catch on.  In the past, fat client software got feature fat because companies (1) rarely performed human factors testing to understand if and why users got confused and (2) there was no automated feedback mechanism.  Marketing wanted more features; the developers saluted and piled them on.

However, the Web has changed that.  It's a worldwide human factors laboratory every day.  By watching user behavior via Web analytics, companies are starting to realize there's a value to making pertinent content and controls available to users -- and value in jettisoning the extraneous stuff, which only confuses users and slows them down.  Because of the Web -- and runaway hits such as the iPod -- the high tech world is finally realizing that intelligent design is a differentiator.

In a busy world overflowing with things to read and things to do, simplification and streamlining -- without dumbing down -- is a feature users yearn for.  With so many things to do, a tool that makes the job easier and less stressful is refreshing.  Happily, companies are finally figuring out what the architect Mies van der Rohe pointed out 50 years ago -- that "less is more."

BEA Buys Plumtree

BEA announced yesterday that it was buying Plumtree Software, a portal software vendor with 400 employees, for approximately $200 million.  This continues a trend in which application and infrastructure vendors are absorbing content management technologies.  For example, on June 14th, Oracle acquired the technology of TripleHop Technologies, a search vendor.  On May 2nd, Progress Software, a database vendor, acquired EasyAsk, another search supplier.  On January 14th, Siebel completed its acquisition of edocs.

As has happened for years in the technology industry, vendors continue to work their way up the food chain.  Infrastructure vendors absorb applications (remember when e-mail wasn't part of the Microsoft operating system?) and application vendors acquire vertical solutions.

IBM's UIMA: "Work Nice" for Search

IBM announced today that it was giving away its UIMA (Unstructrured Information Management Architecture) framework as a way to help search and categorization companies serve the enterprise search market.  It announced that a whole bunch of companies -- including Attensity, ClearForest, Cognos, Endeca, Factiva, Inquira, iPhrase, Inxight, Kana, SAS, SPSS, and Temis -- plan to use the framework.

Given IBM's decades-long penchant for alphabet soup abbreviations (remember JCL, SDLC, SNA?) it's not always easy to figure out what the product does.  For those of you old enough to remember it, UIMA is somewhat akin to IBM's System Network Architecture (SNA), whose heyday was in the 70s and 80s.  While SNA concentrated on bringing order to communications (albeit IBM's version of communications), UIMA concentrates on search.  It defines a framework where a set of analyzers and categorizers can act on a group of content.  What's exciting is that this allows collective technological power to be applied to cracking the meaning of content.

Since many search technologies were initially invented at universities, there has been a long history of search companies announcing that their technology is better than others.  This is often true, but typically for a certain type of search problem.  Put another way, each search company analyzes text in a different way and comes up with different characterizations of the content.  The real power comes from letting multiple analysis engines at the content and then leveraging their different viewpoints.

The power of the UIMA framework is that it allows these different vendors and technologies to work together -- a first within a market that has typically been highly fractured and argumentative.  So IBM is now supplying the enabling technology; it will be interesting to see if the search companies can supply the enabling humility to make it work.