Early this year I followed a talk at Bay CHI from Udi Manber - VP Engineering for Core Search at Google - about The complexity and simplicity of the search experience. To introduce his presentation, Manber gave us this statement: "One tier of search queries that we get in any given day are unique and havent been seen before" and emphasis on the Google's User focused approach to handle these queries.

As we all know the Google search engine simplifies our lifes as much as possible, like using place of the query to make results more accurate, automatically spot and correct typo and spell. It's somehow Google that give us what we need, not what we said. They always assume that we are right, if the result is not what we are looking for then it's their problem and they don't fault somebody else. 

At Google, they have a kind of mantra to illustrate this approach. I didn't catch all of it*, but it sound like that:

The user don't know what he is looking for, it's our problem,
The user don't know how to spell his queries, it's our problem,
...

The advices I retain from a Product Management prospective are that to provide a great user experience we should not consider design as an afterthought... it's the heart of the problem. Start simple, get the maximum data from users, visit them, ask for feedback and why not include a "call home feature" in the product to collect usage statistics (anonymize data and get user agreement first of course).

It's also better to work with progression from simple to complex and add features separatly in order to not overload the product. A product can have tons of features so if the user needs more power then he can get it, but this is also critical to make sure they are not in the way for the average user. This rise a key issue that we have to address: How to make users discover new features? E.g. "tips of the day" or "what's new?".

We should not compromise on simplicity, everything in the product has to be very clear, should need a very short learning curve (who read manual?).

Manber concludes his talk we these advices:

"Best design are build with creativity, intuition and data. The key is to have clear goals, but be flexible how you get there. Data obtain from the web are short term and do not help to define strategic feature. However Get everyone to participate from day 1, Listen to everyone from day 1".

 This talk remind me this slide from Garr Reynold, we should not confuse Simplistic & Simplicity. 

* The podcast for this meeting is not available