Outside of the number one, everyone complains: BARC CEO Sunil Lulla

Industry has always been aware that niche categories try to use the data and navigate the cuts in a small way, says Lulla

The last one week has been a roller coaster for the Broadcast Audience Research Council, or BARC India, the world’s largest audience measurement system, after the Mumbai police claimed to have busted a scam where three channels were paying sample homes to watch their channels. On Thursday, BARC announced it would pause the publication of weekly data for news channels for 8-12 weeks. BARC CEO Sunil Lulla speaks to Vanita Kohli-Khandekar about the controversy. Edited excerpts: 

Q: What do you make of the controversy around ratings and BARC? And why the pause?
 

A: We are like the dholki in the baraat. Whenever there is a baraat, dholki bajti hai (We are like the drum in a wedding, everybody comes and beats us). Outside of the number one everyone complains. Your view of us cannot be based on your rating. I used to be on the founding board of BARC and we realised then that it is very important to move to a place where we can capture the growth of TV. According to global measurement norms, reporting on channels with less than 0.5 per cent share in viewership should not be more than once a month. Industry has always been aware that niche categories try to use the data and navigate the cuts in a small way – say Mumbai city at 9 pm. So niche category protocol and reporting has to change. The techcom (technical committee of 20 people representing all stakeholders) believed that this was an opportune time to reflect and work on this.

Q: Where is the BARC system currently -- size, sample, width, and depth?
 

A: India is at 197 million TV homes and 836 million viewers. We are in 44,000 homes covering 187,000 people. The plan was to go to 55,000 homes (over 225,000 people), but the Broadcast India Survey (the establishment study on which the sample data is based) was delayed due to the pandemic and lockdown. The survey takes one year. Assuming there is no further problem, we should reach 50,000 homes by the end of March 2021 and 55,000 homes by the end of that year. 

Q: Is it adequate to measure viewership in a country with India’s size and heterogeneity?
 

A: I am commonly asked why not more sample homes; because they are sample homes. A sample is a reflection of the population. A drop of blood tells you what you want to know about the body. India is a wide country, there are a number of variables we look at in the sample – statehood, education, location, population strata, NCCS (socio-economic status), signal (cable, DTH, internet), and languages. When you look at languages spoken, very few will say English, so the sample will reflect that. The same principle applies to Hindi, Tamil, Telugu, and other languages. Just making the reporting changes will make things a lot more stable. The broadcaster challenge is to get the best measurement. We have a certificate from the Indian Statistical Institute, Kolkata (certifying the representativeness of BARC’s panel design and household selection) and CESP France (validating the representativeness of BARC’s TV measurement panel). 

Q: What are the checks and balances in place to ensure the sanctity of raw data and final numbers?
 

A: We are statistically built, so we look for outliers in the data. If there are 10 channels in a genre and there is anomaly – increase in time spent at a certain time or at the usual time or new audiences from different clusters come -- we are alerted. Many times, it is a normal thing. For instance, during the lockdown, non-prime time viewing grew, or if a channel has, say, a James Bond film festival, you binge-watch. But if there are unusual spikes, we investigate. The algorithm picks it up; the data scientists look at it. Every week there are a lot of outliers – many legitimate. Then there are errant homes which have been penetrated. Ratings are a currency for the Rs 32,000-crore television ad business, not for content and broadcasting. Advertisers need stability of data. So if things are fluctuating wildly, we investigate. 

A statistical process determines where the boxes are placed. Also, every subscriber (to our data) signs an end-user licence agreement. We have filed police complaints across many states. These are usually local complaints, but these can be filed only if there is evidence. Usually, we get no support from authorities because they (genuinely) have other priorities. We have been trying to criminalise the breach of a panel home, but it is difficult. Infiltration means you are breaking the end-user licence agreement. Plus, replacing a panel home is an expensive process. Who pays for it? Then there is a disciplinary process (on our website), which lists four levels of offence and the punishment for each. This varies from penalties to suspension of rating and termination of membership. So far, there has been action only up to level two.

Q: Why is there this popular perception that ratings are tampered or can be tampered with?
 

A: The perception is among people in the media business. We are measuring the mass and the popular in India, and we give incentives to these homes on norms that have been laid down. These are for the inconvenience, the use of electricity for the meter. We don’t pay them for watching anything. But people figure out how to gamify the system. Isn’t it better to invest in content?

Q: Can BARC choose to only measure genres like entertainment, for which the sample is statistically stable? Can the suspension of ratings for news be permanent? 
 

A: There is a social value to news that cannot be ignored. That is why the revenue share of news is more than 8 per cent (which is its share of total television viewership. Usually revenue share is proportionate to share of viewership).