Advertisers pay big bucks to place their products on popular television shows, and one of the primary ways to rate said popularity is by use of the Nielson ratings. TiVo has a software flaw that prevents these ratings from being accurate.
Nielson uses a method of sampling households to gain a representative figure without knowing exactly what every viewer in the country is watching. Here is how they explain their method:
Nielsen TV families are a cross-section of households from all over the United States. We carefully draw our samples in a way that offers every American household with a television an equal chance of being selected.
When TiVo customers select a program to record, they do so one of two ways – by recording that individual airing or by selecting to record all of that shows episodes using a “Season Pass”. The latter method is what causes the problem with ratings. This is due to a problem with the TiVo software that seems to choose recording times at random when there are multiple options.
For example, let’s say you’re recording “Program A” using Season Pass. This program airs every Monday night at 8:00PM. Sometimes TiVo decides to record it at 8:00PM, but other times it may choose to record the re-run at 10:00PM. It seems to do this randomly, and even if you choose the option to record new episodes only.
In this way TiVo skews the ratings by not properly portraying the intention of the viewer. The viewer probably wants to record the first showing of the episode but that wish isn’t always carried out by their TiVo.
No one seems to dispute Nielson as being the de-facto statistics gathering source for television data, but someone should really investigate the effect this issue has on the numbers.