Interesting - OK, from a bunch of probes we actually do get something useful - even if it comes from left field (Neuromarketing study).   The study, using a fMRI device also showed the CRM questions typically used to measure goal attainment are not effective .. not the right questions to be asking your visitors.

"…The questions related to simple, rational satisfaction (e.g., “How satisfied are you with ____?” and “How likely are you to recommend ____ to a friend?”) produced no enhanced brain activity. The eight customer engagement questions (e.g., “_____ is a brand I can trust.” and “I feel proud to be a ______ customer/user/shopper.”) did produce higher levels of brain activation which correlated with the subjects’ responses to the questions. Fleming also reported that there was a strong correlation between the level of engagement measured by the survey responses and actual purchases."

Aside from the measurement device, measuring visitor engagement requires asking the right questions - one's that elicit an emotional response.

"…. By identifying brain areas that seem to respond to an emotional attachment to the product, this work provides a bit more “marketing mapping” information useful in evaluating subject responses to product designs, advertising messages, etc. "

It may be that, before a series of questions are input to be asked of customers - testing should be done, first, to see what questions your customers typically react emotionally to.

"….Developing engaged customers takes a lot more than clever ads, or even a highly effective marketing plan. Rather, one needs a culture of customer engagement that starts with the product concept and design, flows through marketing and merchandising, and reaches into every aspect of customer interaction and support. While occasionally a company can treat its customers poorly and still engender surprising loyalty, that isn’t the approach we’d recommend. Try and do everything right; you probably won’t be perfect in every area, but a small lapse here or there may not prove fatal if the other areas perform well enough. Don’t count on neuromarketing as a major component of the process; rather, let it be part of the research process that leads to products and marketing that maximize customer emotional engagement with the brand."

What I get out of all of this "Neuromarketing" - Engagement can be measured as an emotional response to a brand experience.  And if you can get a series of questions or even a series of "actions" (IE: navigational experiences of a website) that corresponds with an emotional response from a majority of test subjects - you can then make out an "Engagement Scorecard" based on how visitors answer questions and what actions they take.