Most broadcasters see fault in TAM Media Research's small sample size

NEW DELHI: NDTV's petition against Nielsen, Kantar Media and TAM Media Research has opened a can of worms with a number of broadcasters pointing out flaws in the methodology used by the television audience measurement firm. TV viewership data that is generated by TAM is a critical currency for the media industry that depends on it to make most of its investment decisions.

Most broadcasters allege that the main problem with TAM data is the small sample size. While TV viewing homes in India have more than doubled from 60 million in 2005 to over 130 million in 2012, the sample size being used by TAM (homes which have PeopleMeter installed) has only increased from 5,000 in 2005 to 8,150, 46-47% of India's television viewing homes.

"The sample size is low for a country with a large population and tends to get skewed towards certain demographics and geographies," Lynn de Souza, chairman and chief executive officer at Lintas Media Group says. "There is a need for a larger sample, in terms of size and variety."

She says there are bound to be sampling errors due to the small size. A former senior executive with a television channel, who did not wish to be named, said that if the sample size was large, manipulation would become a difficult task. "A small sample size is more amenable to manipulation and even targeting only a select number of homes can cause a significant change in the data," he said.

While TAM provides information on the viewing patterns, broadcasters get no information on the viewer, his lifestyle, social standing or consumption pattern, elements that are often of great importance to marketers.