I'll echo Neil's statement, that metrics are only valuable if they
meet business goals or industry needs.

The shift from page views to time spent per site does provide a valid
solution for measuring RIAs in an attempt towards equalizing
measurement practices. Yet, the value is somewhat diminished for sites
that seek to provide a rapid web experience. Time spent may be a
valuable measure for media and entertainment sites, yet an online
retailer who can quickly usher a customer through the site resulting
in a conversion doesn't necessarily value that metric. This is similar
to why Google has dropped in the new time spent rankings, because they
seek to expedite the discovery process.

Companies need to evaluate audience measurement metrics based on the
goals of their unique online properties and those metrics may not
always align.

John


--- In webanalytics@yahoogroups.com, "Neil Mason" <nmason@...> wrote:
>
> Hi there,
>
>
>
> It feels like it's almost back to the future on this one...
>
>
>
> Back in 2002 NetRatings acquired a European based audience measurement
> company called NetValue. One of the main metrics NetValue produced (that
> the others didn't) was "Total Duration" ie the total amount of time that
> people spent on a website, on the grounds that this was a better measure
> of user engagement. Their philosophy was that people needed to think
> about online media in the same way as offline media - obviously they
> were a bit ahead of their time....
>
>
>
> I'm not a Jim, Eric or Avinash but this is obviously going to change the
> way that people think about the medium. In one way it may help us move
> towards a measurement standard in our industry as the page view has
> always been a bit of a dodgy metric. There will still be reconciliation
> problems between audience panels and web analytic systems as total
> duration will still be dependent of the definition of a session length
> etc etc. At the end of the day when you are looking for standards, it's
> inevitable that you will always be playing to the lowest common
> denominator and so it's going to be sub-optimal. What's right for one
> industry is unlikely to be right for another.
>
>
>
> Cheers
>
>
>
> Neil
>
>
>
> ________________________________
>
> From: webanalytics@yahoogroups.com [mailto:webanalytics@yahoogroups.com]
> On Behalf Of romanojon
> Sent: 10 July 2007 19:59
> To: webanalytics@yahoogroups.com
> Subject: [webanalytics] Nielsen NetRatings Switch-Off to Time on Site
>
>
>
> Hey everyone,
>
> I just read a PDF located at the following address:
>
> http://www.nielsen-netratings.com/pr/pr_070710.pdf
> <http://www.nielsen-netratings.com/pr/pr_070710.pdf>
>
> From what I gather, it seems that due to the increasing pervasiveness
> of applications based in AJAX and other media driven applications
> contained within a single page, Nielsen has taken to using the primary
> metric of Time on Site.
>
> Does anyone have any opinions on this or can anyone tell me how this
> affects them? For us, not having any relationship with NetRatings,
> and very limited involvement with what people would consider Web 2.0,
> I have no perspective on the magnitude of this announcement. Judah
> always preaches 'Context' and I agree with him on that. So, having
> said that, is this transition slightly premature in that it only
> applies as a major measurement of top sites in addition to their
> traffic for engagement, or is it something so big that I just can't
> conceive of it yet?
>
> I realize there is virtue and value to measuring the attention span of
> the user. I see it as a margin to shoot for in terms of presenting
> useful options on the site. In the event that we would add numerous
> videos or instructional content, I would expect its useful in knowing
> how long people like their videos to keep feeding them things which
> conform to their expectations. Beyond that, and as for using as a key
> metric on which to base performances, I still see it as subordinate.
>
> Maybe Jim, Eric or Avinash have some insight on this or can lend some
> wisdom here?
>
> Sincerely,
>
> Daniel W. Shields
>
> http://danalytics.blogspot.com <http://danalytics.blogspot.com>
>
>
>
>
>
> [Non-text portions of this message have been removed]
>