Wider sample base, more transparency needed

Television is the second largest advertising medium in the country - more that Rs 13,000 crore rode on it last year - and ratings data is the measure of accounting for the industry. As the Broadcast Audience Research Council, the audience research measurement body formed by broadcasters, advertisers and advertising agencies, gets ready to put in place a new television measurement mechanism next year, we ask the various stakeholders what they expect from the revamped system and the features they would like included to make it a more robust currency than the current one

Given that audience measurement systems don't operate with unlimited resources, the call here is not really for a wish-list but more for managing expectations of various stakeholders and eventually to have a system that will do the job well. Despite the ever-changing media landscape, television continues to be a mainstay medium and an efficient audience measurement system will always be integral to successfully execute our communication plans.

Among the many things that make for an efficient audience measurement system, the first is the issue of representation. The Indian television scenario has undergone many changes and will continue to do so. As media agencies and marketers, we welcome this dynamism, and, in fact, thrive on it. We wholeheartedly embrace the vast range of viewing options, habits and genres that exist today on Indian television.

To ensure we have reached out effectively and meted out justice to our stakeholders, an ideal audience measurement system should be one that is both representative and inclusive - demographically and geographically. Simply put, we do not want our young planners and buyers to be pre-conditioned by the television ratings they see; we would like the measurement system to be robust and representative enough to capture any evolving viewer dynamic. This is not just about television - and channel - performance, but it is really about us; being always armed with data that questions what we think we know and challenges the status quo.

I have been using research long enough to appreciate that no system will explain the presumed inconsistencies and gut feeling that we may have about trends on television viewing. Therefore, for me a second key criterion is the availability of - and access to - the experienced research minds associated with the audience measurement system provider; they would not just help put to rest our concerns but also be the watchdogs for a better tomorrow.

Audience measurement systems provide a universal currency and as such tend to operate in monopolistic conditions. There is, therefore, an enhanced need for self-vigilance and an ever-evolving, ever-questioning process that justifies its stature as the industry benchmark.

This brings me to the last, but by no means the least, relevant criterion - transparency. This would mean the process of checks and controls that are regularly reviewed and refined by the industry to ensure that all the stakeholders are committed to the system. All kinds of research have their limitations but it is of utmost importance that we keep the faith. There is no substitute for efficient measurement; we don't just want it, we need it.

A final point of view: Audience measurement systems have a lot more to offer than just weekly ratings. The system is a treasure trove of insights for brand and channel planning. Let its existence be optimised manifold; let us spend more quality time in here.

Jasmin Sohrabji
CEO, Omnicom Media Group

 
As a broadcaster and distributor, my primary concern with audience research measurements today is that the data is suspect. The credibility is low, made evident by not just subtle but some very obvious lapses. When the current audience research system, which the industry uses, shows IPL ratings that are lower than last year's, despite indications to the contrary, a subscriber is bound to question its veracity. It gives a misleading picture to advertisers and also to us. After all, when taking a call on buying rights to programmes to air, we assess past ratings as well.

The first biggest drawback is the sample size. A small sample size, which we have seen in the present system, makes the entire extrapolation weak. For a country like India, with its many nuances such as cable-dark areas, single TV per households, different social strata present in the same neighbourhood, capturing a cross-section is a challenge. In particular, in the current system, the underlying establishment study is not transparent. Hence, we usually have no idea on the differences that get captured.

A large and well-chosen sample size will make such measurement more representative and far more accurate. The sample size needs to be about three times that of the current sample, and that will mean covering at least 20,000 to 25,000 households. The placement of the boxes must be done in households, which are chosen scientifically, since they will be the basis of the projections. Only then the sanctity and clarity of the statistics will be guaranteed.

The second most crucial aspect to make audience measurement systems more fair and robust is introducing separate teams for various processes that eventually give us the data.

Make no mistake. Separating the processes of installation of the boxes, collection of the data captured, and a thorough analysis of the data is vital for a robust audience measurement system. How, then, can the same team perform all the three tasks? Separate teams should take up these three jobs with no correlation between them. Otherwise, fudging and stretching of the data from the coveted black box is unavoidable. The objectivity of the entire system suffers. This 'separation' will bring in more checks and balances.

Yes, this step will be difficult to implement - much like a democracy that finds it tough to maintain the separation of its legislature, executive and judiciary. But it simply has to be done. Compromising on this step could jeopardise the credibility of the data that finally reaches the stakeholders.

The third measure has to be on-ground monitoring. The current system claims to do so but the end results show that there is a severe disconnect between the ground reality and what the system claims. A more stringent on-ground monitoring will ensure the boxes are secure in that they are not tampered with, lie idle or randomly shifted elsewhere. A large sample size will prevent this and help monitoring.

The need of the hour is to maintain the integrity of the data. This way the validity of the data will also go up. BARC will ensure that an active monitoring programme is in place to prevent tampering.

Man Jit Singh
CEO, Multi Screen Media Pvt Ltd & President, Indian Broadcasting Foundation

For advertisers owning brands in the business-to-consumer space, it is very important to draw up a media plan that is dynamic. Canon is one of them and we need our media plans to be able to swiftly adjust to any changes in TV viewership. We cannot afford plans that would take a week to be tweaked, or which will bear fruit only in a month's time. After all, we spend 70 per cent of our above-the-line media budget on TV. Hence, our interest in TV viewership measurements is substantial.

As a B2C advertiser, one of the first criteria for a robust TV audience measurement currency is that it is representative of a wide cross-section of the viewership. Keeping the Indian TV viewer in mind, we need to have fair integration of true behaviour of audiences from the lower tier towns too. We will not be content with narrow measurement restricted to viewership in metros only. India being so diverse, the sample needs to be representative of our consumptions. Accurate, relevant and meaningful audience measurement instrument should provide a window to the patterns in smaller towns across regions, and even across the different strata in a given location. A sample drawn from not just SEC A and B but also C, for example, will be more representative and useful to us. It is foolhardy to paint Indian TV audiences with a homogeneous brush.

The second criteria is perhaps more subjective but nonetheless much needed in TV audience measurement systems. While other media have remained more or less same in their audience patterns, TV is increasingly showing wide variations in viewership trends with digitisation in place. Moreover, for brands which talk to the youth, TV viewership has become even more complex. Technology has ensured that the youth consumes media from two different screens even as they watch TV. There might be a programme on, but they will also be busy on their handheld mobile devices. The measurements should reflect this change.

The third thing I feel a great need for in TV audience measurement systems, is the equal need for speed as for accuracy. Granted, that surgical precision in such measurements is something we should aspire for. But equally important is the need for real-time updates. In six weeks, if we end up spending Rs 10 crore, then that would mean our media plan spends no less than Rs 2 lakh per day. With that kind of a sum, I better know the effectiveness of the allocation I do daily, rather than have a weekly snapshot.

This would need us to be on our feet to change a whole media plan in a week's time with tracking on a daily basis. However, for this to yield results, our media buying patterns should also change, rather than just audience measurement. Advertisers could buy a basic pack of spots from channels, say general entertainment, and then top up spot purchase depending on the daily fluctuations as evidenced by a more robust audience measurement system. The clients need to rethink and get their agencies aligned for this to happen.

With a fragmented TV viewership universe, it is time to have a whole new measurement for it.

Alok Bharadwaj
Executive Vice-President, Canon