By Dennis D. McDonald

Steve Rubel's The Imminent Demise of the Page View is yet another warning of the eventual demise of the "page view" as a way for measuring ... what?

I know as do most folks that page views have a bad reputation when it comes to REALLY measuring the interaction between users and the business of a web site.  It's not just the gradual increase of within-page interactions (via AJAX etc.) that don't get measured by page views, it's all the other important stuff that goes unmeasured such as willingness-to-return, RSS use, willingness -to-reference, etc.

Every week I base a post here on the previous week's top ten posts measured by page hits. I know looking at my data that using hits as a measure of usage has significant drawbacks. Most profoundly, page views give no idea of the level of engagement of the user, as I've already written in Page Views Are Dead! Long Live Page Views! (A Discussion of Page View Alternatives).

ALL KIND FOOD is not advertising-driven. This blog is, pardon the expression, mostly a labor of love. I use it as an advertisement for my consulting services, as a "live" extension of my resume, and as a conduit for my professional networking activities. Whether 1 person, 5 people, or ten people visit a page or its representation in a feed reader is not the real determinant of my happiness. I care more about who visits the pages and then  does something to follow up -- by a comment, by email, or by a phone call.

In other words, I'm more interested in whether or my web page becomes a step in the communication between me and the reader, regardless of how that communication takes place. I'm well aware that page views, as are most measures, distant proxies for measuring that type of communication.

Perhaps it is inevitable that web usage metrics will become more like TV viewing metrics as it becomes harder and harder to use page views as metrics for use or user engagement. By that I mean that a separate process will be used to interactively measure on an explicit basis what people are actually looking at or doing at a given time. Maybe that's what the individual advertising metrics firms are already developing as they roll out technology to measure AJAX based interactions.  Again, that's not something that concerns me directly given my personal lack of advertising use.

As I discovered while consulting with call centers,  the single most important detail I wanted to know to help me interpret incoming call statistics was why people called. If I could get that data element attached to a call transaction record for downstream analysis purposes, I felt much better, since the reason for a call drives satisfaction or dissatisfaction with the outcome of a call transaction. (And in relation to this, we need to remember that satisfaction or dissatisfaction with the quality of a transaction is not necessarily correlated with the satisfaction or dissatisfaction with the outcome of that transaction.)

Whether or not cookies can be used to classify the purpose of an incoming web page transaction I don't know. I do know that most of the statistics that Google Analytics records about visitors to All Kind Food are really by-products of transaction logistics and have nothing to do with the identity of the visitor; I really don't care about the visitor's screen resolution or operating system though I realize that for some folks these items will be critical.

I'll go back to ideas I've already discussed; perhaps an ideal system would be one where folks who are willing to do so would voluntarily broadcast details of themselves when they surf the web. For example, I might be willing to publicize a bit about myself when searching or scanning my feeds if I thought doing so might facilitate a more targeted and personalized transaction. I am well aware that Google may already have the ability to do all this by tying together the details of my surfing and other Google related behaviors, but I haven't yet figured out for myself how searching Google while signed in benefits me more than searching Google while signed out; perhaps someone here can illuminate me on this question?.