This monitoring change could dramatically alter how network programmers decide on what programs get and stay on the air, since it allows much more comprehensive breakdowns of viewer demographics than currently. Under this new system, a supposedly low-rated show like the recently cancelled Arrested Development will still be on the air, because with the much higher sampling rate Nielsen could note that there is a very loyal viewership for this show that is highly desired by advertisers. Also, with 50,000 active meters there will be no more overnight initial ratings since the ratings data will now come in faster.
It would be very interesting to see the demographic breakdown of The CBS Evening News under the new system--is CBS getting the right viewer demographic to justify paying Couric US$15 million per year?
Sample expansion is certainly a good thing but we are most interested in some other refinements. Nielsen is gradually moving to Minute-by-Minute data. They have been reporting "snapshots" of what was going on during the mid-minute of the quarter hour. Lots of room for error using that approach. With minute-by-minute, you get more accurate estimates of audience flow, source, destination and such i.e. where did the viewer come from, where did they go.
Years ago, Nielsen was working on facial recognition in order to automate the system...too complicated. What they were REALLY trying to get at was a measure of Attention Paid. Entities investing in adspace want to get in front of people actually paying attention and what good is/are dollars spent on putting a commercial on the air if people just happen to be in the room and not really paying full attention to what is on?
The future is in People Meters that pick up and report embedded audio signals from encoded sources that are beyond human hearing range. Diaries and set top meters will be history.