Information architects and business strategists can't afford to work in ivory towers or be limited by narrow departmental perspectives.

The authors of Strategy Safari proclaim, "We are the blind people and strategy formation is our elephant. Since no one has had the vision to see the entire beast, everyone has grabbed hold of some part or other and 'railed on in utter ignorance' about the rest." Swap "strategy formation" with "information architecture" and you've just described many of the heated debates at our conferences and on our discussion lists.

And I like the 'experiment' to show where people come from to read the interview, it's right up my street.

Posted by: Mac on October 24, 2002 04:40 AM

---

 

Confession & Retraction:

I asked a friend of mine at HP why they don't map "Journada" onto "Jornada." Her response:

"Because you're the only one who misspells it "journada". ;) Really, that's the reason. The frequency of that particular misspelling is very low. We base the correction on actual mistakes made as it takes processing time to check all variations and the speed of search is a key factor in user perception of its usefulness."

Good answer! Sorry HP! I'll pick on Microsoft next time :-)

Posted by: Peter Morville on October 25, 2002 12:42 PM

---

 

Peter, please retract your retraction.

> "Because you're the only one who misspells it "journada".

No you're not. If you search for "Journada" on google images, or google web you will see that your friend is telling you fibs. And even if they weren't, it's still wrong to dismiss your comment just because you are one person. You wouldn't be the only person to misspell it, you would be the first person. If HP did something about the first person, then there wouldn't be a second person.

I treat every 'failed' search as an opportunity to improve my intranet search. If checking for alternates is a factor in the speed of the HP search then they need to re-design their search engine. That's like saying you don't allow searches with Z in the words because the Z key on your keyboard doesn't work (bad example, but it's the best I can come up with in 20 seconds).

I have just tried to search for "Journada" on the HP site (US) I get two results, whilst on the UK site I get one result. This means that someone in HP has put the meta tag "Journada" in their keywords list to try and solve your personal problem. However this is almost worst than getting no results, as a user may think that HP understand that "Journada" = "Jornada" and that these are the only results available.

I see a site search as being rather like a flare gun. When I perform a search I am getting a glimpse of the structures and vistas of the site I am searching for a split second. A flare that only show me a small disjointed view is much less useful that one that shows me a glimpse, however brief, of the bones of the site.

And while I'm on the subject. When I search google for "information visualisation" I don't want it to say Did you mean:"information visualization", I want it to take the results for both searches and combine them into one set of results for me. Whenever google says "Did you mean... it's highlighting a failure of the search engine and it's trying to show off and illustrate how clever it is instead of giving me what I want. But that doesn't mean that I want it automatically jumping to another set of wrong results automatically as it seems to do sometime.

Ahhhhh. Glad I've got that off my chest.

Posted by: Mac on October 28, 2002 03:13 PM

---

 

Sorry, but it's my policy to never retract retractions...I'm too afraid of having to retract retracted retractions. But I am glad you made the case for handling Journada...there is a cost/benefit tradeoff here and ultimately I agree that the marginal cost of handling another equivalent term should be pretty low...

Posted by: Peter Morville on October 28, 2002 04:15 PM

---

 

What if you detracted a rectracted detraction instead?

Posted by: Mac on October 29, 2002 03:38 AM

---