In the physical world great salespeople ask questions, listen, then respond with suggestions that guide the prospect toward making a purchase.

Since Web visitors are reluctant to answer these questions in forms, we need to use more subtle content-based ways to ask about their interests and needs.

The personalization systems that tried to guess what a Web visitor should see were too expensive and too complex to have a positive ROI. That’s why most of those vendors no longer exist.

It turns out that it’s rather easy to apply a small amount of personalization to a Web site, which most Web visitors seem to appreciate.

“I was in athletic clothing, I was even perspiring a little”

What if I’m a beverage saleperson and I offered you a bottle of water? I’d say there’s a good chance you’d want it. Especially since ice cream can make you thirsty. If didn;t want water I would guess that the majority of people who were offered the water in athletic clothing and perspiring would take it. I could be wrong, but I might offer 100 waters and see what the results are before I make a conclusion. I might even need to offer 1000. But at some point I might know for sure. Numbers don’t lie.

As far as control it is a great strategy especially for content but vistors are always in control in a retail environment (unless their mouse freezes). Many times presenting more than one choice or pathway as you call it confuses users by presenting too many considerations. The “paradox of choice” effect if you will. How do I know? Data and testing.

I guess the overall issue is personalizing based on personas as you mentioned in the diamond example. I’ve always maintained that personalizing to high-impact segments based on actions is better. Personalizing at the user level as we found out in the late 90’s and early 00’s is probably not scalable (or effective) for anyone not named Amazon or Google.

When a site is redesigned to be more personalized several things change.  Page views, especially for some pages, may go down drastically, due to Ajax and Navigation changes – at least, that's what I've noticed.Another sticking point – Anal…

By “observable behavior” I was referring to data generated by visitor clicks on a website. That raw data can tell you what people do, but not why they do it.

You CAN gain insight from observable behavior as long as you have context to understand it.

Persuasion Architecture uses personas to give context to those visitor clicks. Personas give you insight into not only what people do, but why they do it. Personas allow you to understand motivation behind behavior, which is what really matters if you want to have true customer insight.

Personas allow you to plan and measure those viistor clicks and assign context to better understand that behavior.