Had no idea that TV Commercials were never measured for audience effectiveness - Seth Godin says they haven’t been and both the Networks and the Advertisers don’t want commercials measured that way.

"Measurement is always tricky, because people believe what they want to believe and find the numbers to back it up. In both cases, we’re seeing how advertisers and media companies are complicit at weaving a story that doesn’t really hold up. How many "hits" did your web page get last week? And what, exactly, does that mean?"

In my metrics work at corporations - my clients often what to know the effectiveness of a page and the "engagement level".  What metrics I have come up for that?

Next Clicks - the total amount of pageviews (or visits) off the landing page is one measure of "engagement" and when divided by the total traffic to a page gives you the % that wished to click through.  However, I’m coming to believe that "next clicks" ought to be defined for the action that you want someone to take once the reach your site.  

For example, if they land on your homepage and you want them to click on a link to buy a CD - then next clicks would be the percentage of traffic on the page that clicked on the link (ie: if 7% of the traffic clicked on the CD link, that’s a 7% engagement - of the traffic of on the page).

% Next Clicks / % of Unique Visitors to the Landing page  - then click through the content of your desired action page - I came up with this metric to define how often new visitors, as a measure of all visitors are more engaged in the content you want them to be. 

For example, if 7% of the traffic clicked through to the CD page, but 5% of the 7% were uniques - it’s fair to say your offer is being shown to mostly new people who haven’t seen it before - and if they buy it then - it’s impulse buying probably.   If, on the other hand, 6% of the 7% of the  traffic from the landing page reached the CD page and was return visitors (and they don’t count as uniques as they’re counted the first time they arrive) then you might say they are making a "considered purchase".

Time spend on action page - How much time do they spend, on average, on the page where they are sopposed to take action.

How does this apply to commercials?  here you’d have to first define the target audience and then figure out if the Commerical had the desired effect - "brand awareness" or a particular promotion (ie: buy new diet pepsi).  The problem is - it’s really hard to tie a commerical to going to the store and buying the pepsi - and maybe the only way to do it is to look at the consumption after the commerical was run in target areas where the commerical is run - and that’s not an easy thing to do as there may be delays in getting the information and distributors may deliver (ie: soda) several months before it’s actually consumed.

In short, metrics for commericals are probably very hard to set up and that’s why they never have been - till now. It will be interesting to see how Nielsen rates the viewership of commercials.

"Second, Doug Karr writes in to take newspapers and the Audit Bureau of Circulation to task for the changing standards in newspaper ratings. The numbers are a lot less strict… and a lot softer… than they were a decade or two ago, he reports."

As data gets harder (I know this sounds strange) to get a hold of (fuzzier is a better word) due partly to market fragmentation media outlets are probably cooking the numbers when they can’t get solid information.  Many numbers are soft - they depend on how you decide to calculate a metric.