Following the I&B ministry’s order to share raw-level data with broadcasters in a transparent and equitable manner on August 10, it is reliably learnt that BARC has proposed to share respondent-level data instead.
This tweak given by BARC with respect to the ministry’s standing instruction allegedly goes against I&B Secretary Apurva Chandra’s statement that “raw-level data will be very useful in checking manipulation”.
Raw-level data is what the peoplemeters report from households. Raw-level data is unprocessed data, which means this data is without outlier exclusion.
Respondent-level data is the processed data after running outlier management tools and algorithms, which is, in layman’s language, the final data that we see today as TRPs.
This respondent-level data is both on individual and household levels.
Both raw-level data and respondent-level data have a similar acronym – RLD.
According to the copy of the ministry’s order, BARC had requested permission to share audience measurement raw-level data (RLD) with broadcasters.
BARC informed the ministry that RLD data has been sufficiently masked and anonymised and does not include personally identifiable information of panels thereby ensuring the secrecy, privacy and integrity of the panel homes.
BARC also conveyed that they were already sharing the RLD data with advertisers and media agencies.
With this understanding, the ministry ordered BARC to share Raw Level Data up to the preceding week (i.e. RLD data of the viewership for W-1th week when the TRP is being reported for the Wth week) with the broadcasters in a transparent and equitable manner while ensuring that the integrity and confidentiality of the BARC panels is maintained at all times, the order read.
Further, the ministry asked BARC to must adhere to all the requisites and prescriptions of the “Policy Guidelines for Television Rating Agencies in lndia".
But there was a catch in the BARC’s submission to the ministry.
BARC had said that it was already sharing RLD with the advertisers and media agencies, which is respondent-level data. The ratings body did not clarify it then whether it was unprocessed (raw-level) or processed (respondent-level) data.
Now last week, the ratings body in a detailed presentation to the ministry conveyed that it was sharing respondent-level data with the advertisers and media agencies as they did not need the raw-level data for forensic purposes. Instead, they needed the data on more granular levels for their media plans.
However, BARC told the ministry that it can share the raw-level data if the ministry feels otherwise.
Owing to the ongoing four-week rolling average data released for the news and niche genres, BARC is apparently telling the stakeholders and the ministry that it would share the rolled data with the news broadcasters.
This essentially means that it would be of no use when it comes to checking manipulations and would offer nothing to the news broadcasters as respondent-level data on a rolling basis is as good as the current ratings released on a four-week rolling average.
For example, while decoding the data with broadcasters, it did not give clear picture if Chandrayaan-3 coverage garnered more gross GRPs than the coverage of the death of Yevgeny Prigozhin, the bombastic head of Russia’s infamous Wagner group.
It may be noted that the Chandrayaan-3 landing on the lunar surface on September 23 broke all the viewership records on digital proving that everyone in the country was glued to their screens to witness the historic moment.
Everyone is curious to know whether such an event and the build-up around it delivered historic gross GRPs on television.
When BARC has already put a caveat that it will be minus one week, then it is not clear why it is resisting the unrolling of the RLD and hiding the true picture.
The ministry was under the impression that the raw-level data is the unprocessed data without outlier exclusion which will give an idea to the broadcasters if their final TRPs are justifiably treated in comparison to others.
It may be noted that the manipulation in processing the raw-level data came to light in the audit report of Acquisory Consulting.
Basis a chart from the audit report, BestMediaInfo.com reported how the viewership data of Times Now was reduced drastically from raw to final data to keep Republic No. 1.
Not only this, Republic’s numbers were not touched in the weeks between July to October 2017. It was the same period when Goswami had chats with Dasgupta from which it could be concluded that his channel was present on the landing page of Kolkata’s largest MSO ICNCL.
In this context, the I&B Secretary told BestMediaInfo.com that sharing raw-level data with broadcasters should be very useful to bring about transparency and check manipulation.
It is yet to be seen if the I&B Secretary will approve BARC’s offer next week which is nowhere close to his objective of checking manipulation through raw-level data.