Serbian authorities claim Nova S ratings low

NEWS 25.04.201916:07
United media

Counting TV viewers has become another statistical discipline frequently present in the media as of late, Belgrade weekly Vreme said in its latest issue.

The use of only the viewer data which suits a certain TV station has been done over the past 15 years since the Nielsen group has been regularly gauging viewership on its peoplemeter but since the Nova S TV, owned by United Media which has more than 20 stations in its portfolio including N1 and all Sport Klub channels, the impression is that both that station‘s competitors and the authorities want to diminish its ratings, the weekly said.

„Ever since Nova S started airing one of the most popular shows in Serbian air space – An Evening With Ivan Ivanovic – and talk show host Olja Beckovic‘s Impression of the Week (Utisak Nedelje), the odium that accompanied Beckovic and Ivanovic has moved to Nova S,“ it added. Both those talks shows were known for their critical attitudes towards the authorities before they were taken off the air, allegedly under pressure from the authorities.

According to the weekly, it‘s clear that Nova S, being only a cable station, can‘t compete with any of the stations with national coverage but that does not seem to be enough: „They have to prove that no one wants to nor should they watch that programme because they are traitors“.

And so, a war of press releases started in which TV Pink brags that the Nielsen ratings data shows that its Hit Tweet show, which is aired at the same time as Beckovic‘s Impression of the Week, has a far greater number of viewers.

„On the other hand, United Media published its statistics which showed that Nova S has the most viewers in Serbia when Beckovic and Ivanovic are on the air,“ Vreme said.

United Media CEO Aleksandar Subotic told the weekly that the Nielsen peoplemeter use a sample on an entire territory which should represent all the households in the country which own TV sets.

„In Serbia that sample is 880 households, including households whose TV signals come via one of the pay TV platforms, cable, IPTV or satellite platforms as well as those who use antennas and do not have pay TV. Also, those households are spaced in line with geographic distribution, socio-demographic distribution, family traits, etc. Devices in those households register how many members of the household watch which channel and when. In Serbia, those 880 households have about 2,500 individual members. Based on the behaviour of those 2,500 individuals, an assessment is made of the viewership among the people who own TV sets in Serbia. I am not sure about the exact number but I assume that they number about seven million because the ratings do not include children under the age of four,“ Subotic said.

„This includes an acceptable margin of error only if the sample is ideal which we don‘t know in the case of Nielsen in Serbia. The sample mush be an ideal representation of all forms of TV viewership, cable, Internet or terrestrial with adequate representation for each system, it has to be a realistic representation of the relationship between the urban and non-urban populations, etc. It has to be completely anonymous because the results will be completely different if just 100 peoplemeters do not reflect true viewership (for any reason). This kind of ratings measurement is open to a large statistical error and sample corruption which leads to incorrect results and is heritage of the past when interactive technology was not available,“ the United Media CEO said.

„SBB uses the services of TVbeat when measuring viewership using data from STBs (set top boxes). When we talk about channels available only on platfroms, the statistical error of the peoplemeter methodology is very high as a direct consequence of the small sample. That is also the reason why TV operators use their own data. In the case of SBB, the data in Serbia is taken from 450,00 STBs in 450,000 households and that number rises daily as users of any SBB pay TV service move to digital technology (D3 or EON). SBB has almost a million subscribers in Serbia and the data from 450,000 households is a representation of the behavior of a million subscribers (households) which is about 40 percent of all households in Serbia. It is completely logical for the data from the high number of STBs is far more precise than the peoplemeter technology data as an indicator of the behaviour of viewers of pay TV operators, in this case SBB. This methodology, unlike the peoplemeter technology, reflects viewership in households, not among individuals and, because of the size of the sample, has virtually no margin of statistical error,“ Aleksandra Subotic told Vreme magazine.