Is our TV measurement future-ready?

19 Dec,2012


By Bipin Mundhwa and Amit Nevrekar


TV audience measurement through the ‘peoplemeter’ is accepted as the most appropriate technique among others for measuring TV audiences. However, the capability of the peoplemeter (an electronic device) is completely dependent upon external factors such as continuous human intervention (Manual Inputs). Viewers need to press their respective button on the peoplemeter remote (as assigned to them) to let the system capture the data on who is viewing what. Which means it is safe to say that it does not automatically capture the viewers and their viewing patterns. It requires the viewer to keep in mind that they have to press the respective button every time they leave the room and enter the room.


In an age, where people forget to turn off the switches while not occupying the room or even turn off the water taps after use, do you think people will remember religiously to switch on and off their respective people-meter remote button every time they enter and leave the room?


There are high chances that the viewer may get interrupted by something and leave the TV room or for a while take an exit to do some other household activities without pressing the exit button. In such situations, the people-meter does not stop capturing the viewership of the audience, even though he/she is no longer viewing. What could be more erroneous than this?


Jaldi 5 with Bipin Mundhwa & Amit Nevrekar: No one influenced us to write book
Bipin Mundhwa

01. It’s interesting that your book is published around the time when TAM is being needled by many stakeholders. In fact it’s also being sued. Is it just a coincidence or planned… given that you’ve highlighted the T, A and M in your book title ‘The Advertising Mess’?

Frankly speaking, we had planned the date for our book launch much before the announcement of non-reporting of viewership data by TAM. The main objective of our book was to provide the necessary practical literature on The Audience Measurement systems in India. Over the years, we have seen different methodologies for the same medium in the same period resulting in different findings and thereby creating a Mess in media buying and planning.


Amit Nevrekar

02. Having worked with TAM as also with other media organizations, why is it that you didn’t raise some of the issues earlier?

Working with TAM was always a pleasure. And we constantly discussed many improvements as part of our professional duty.


There may be people who may say that you have written this at the behest of someone wanting to take on TAM?

Firstly, we have not condemned a particular organization or criticized upfront. The book is about ‘bridging the gap for accountable MarCom’, where we have presented the ‘Bridge’ model which provides a unified B-Score across mediums in a single metric based on engagement levels, rather than inflated exposures. No one can influence us for taking such a mammoth effort whilst working, especially in a media agency, media house with hectic working hours. It’s only our self-inspiration and passion for our industry.


03. Given that there will always be television channels who feel aggrieved by any ratings, there will be people who will raise objections against any measurement system?

Every media professional subscribing to the respective measurement system’s data has the right to raise questions to understand and utilize the data in most logical and appropriate manner.


04. You’ve touched upon corruption. And you say that with digitization, it could only increase?

Yes, we have identified some loopholes in the system which can simplify corruption post digitization and needs to be addressed for robust data representation.


05. One of the primary motives of the book is to talk about the ‘Bridge’ model that both you have developed. Have you had any discussions with BARC about it?

The bridge model is purely developed by us. We haven’t had any discussions with BARC as an organization but we certainly had series of discussions with members of BARC, MRUC and other media veterans of the industry.


In the current peoplemeter, if the TV is ON but none of the panel members have logged in the system (that is, pressed the peoplemeter button), then the peoplemeter would prompt for an input to register the viewing occasion for the respective panel member.


But what if the panel member has logged in the system and is not present in the room. Is there any kind of validation thought about?


However, today’s technological advancement allows us to make every impossible thing ‘possible’. There are Pyroelectric (“Passive”) InfraRed sensors that can detect whether a human has moved in or out of the sensors range. And undoubtedly, these are very inexpensive chips.


Another such technology is used by Microsoft in their Xbox 360 game console, ‘Kinect”, which is a motion sensing input device capable of facial recognition along with full-body 3D motion capture. Which means if incorporated in the peoplemeter, will not only sense the individuals present in the room, but also identify if they are viewing the TV along with their engagement through facial expressions.


Also, a very interesting point crops up, which ideally should have been raised by MarCom professionals. Bear with us as we elaborate.


With the new Digital Video Recorder now commonly available along with most of the DTH providers, individuals have now got into the habit of not watching programmes live, but recording their programmes of choice and watching at their convenience.


For example, if a working individual chooses to record a programme which was telecast from 4 – 5 pm, reaches home at 8 pm, and watches this recorded programme from 9 – 10 pm.




The critical point we are making is, most people with DVR facility have now got into the habit of recording almost all programmes they wish to view, and while viewing the recorded programmes, most of them would forward and skip all the ads.


The question is, at 9 pm in the case of the above example, what will the peoplemeter capture as viewership?


Will it capture the 4 – 5 pm show which is being watched between 9 – 10 pm or will it capture the 9 – 10 pm show which is being recorded for future viewing at the same time?


And will the peoplemeter be able to gauge whether the ads have been skipped by the viewer while watching the recorded programme, which is bound to happen in most cases?




Thus this DVR facility has the potential of making the peoplemeter defunct and useless.


Definitely, the currencies such as TAM should be keeping up with these advances in technology. We are very sure they have an answer for the questions we have raised above.


BUT, we are also sure the industry and MarCom professionals would love to see a practical demonstration, if they claim that the people-meter can take care of these issues. Speaking is so much easier than doing.


According to research paper published by Andrew Green (2010), the people-meter is only capable of capturing what is being displayed on the TV screen and not the actual behaviour of viewers.


Very interestingly, TV broadcasting would be going digital in coming days. It has already gone digital in the four metro cities of India as on November 2012.


In future, there could be options of pay v/s free channels, where the pay channel will have only programme content and no advertisements, similar to the current HD channels.


DVR (Digital Video Recording) technology is also expected to dominate in the near future, which means there would be more flexible and enhanced TV viewership. This would also include storing of favourite TV programmes by viewers who would watch them at their convenience while skipping the commercial breaks.


There is also an increasing trend of VOD (Video on Demand) in direct-to-home households. Would this mean lower viewership for repeat telecast of TV content? Because as the name suggests, VOD is TV content that can be viewed at anytime as per the convenience of the individual viewer. More sob stories for MarCom due to digitization.


The new-age television viewership will include CAS (set top box), as well as others like DTH, DVR, IPTV/VOD, internet TV, TV programmes via YouTube and mobile TV. Conventional wisdom says that subscribers of these technologies will have vastly differing lifestyle habits. As we say, “necessity is the mother of all invention”. For the first time, the audience will be in the driver’s seat for choosing their option of technology of reception. So is the case with TV channels that they wish to subscribe to.


Now, looking at the traditional TV measurement system, the viewership base will be determined by subscription packages, as now the option of choosing a specific channel is available to the viewer. In the recently phased-out analogue system in 4 metros, for a fixed monthly subscription fee, subscribers got access to ALL TV channels.


This means the sampling frame for estimating TV viewership is based on the audience pool that recently had access to all the channels for a fixed subscription fee.


Whereas, the new technology of digitalized reception will make the subscriber’s pool asymmetric for sampling. There are bright chances that the universe (the channels which an individual subscribes to), will fluctuate depending on basis of favourite channel/popular programmes/sports seasons or a new series, as people may change their channel subscription from time to time.


For example, let’s consider a household which has recently switched to the DTH system from the old analogue system. Now this particular household has not subscribed to a particular GEC channel because they felt it was not worth paying an extra fee as they were casual viewers of only one programme on this channel, as against other GEC channels where they watched multiple programmes.


Now after a couple of months, this household again decides to subscribe to that GEC channel they had abandoned because of a new programme of interest to them, along with a sports-pack for the upcoming IPL cricket season. Post the IPL season, they again unsubscribe the sports-pack.


This means the channel’s universe has now changed beyond recognition. This will keep happening as now it will be directly related to pay for what you watch.


In such situation of switching subscription plan from analogue to digital, peoplemeter may capture the audience movement from analogue to digital. However, within digital, the viewership for channels may fluctuate significantly (as these are sample surveys, where 1 panel member’s viewership is extrapolated to around 20,000 audience or more). This could result into vague and half-hearted understanding of channel viewership as practitioners would not know the reason for volatility in channel viewership i.e. Subscribed but NOT viewing OR Channel NOT subscribed.


Until now in analogue mode, the same set of channels that a household with a people-meter had was also the same set of channels that all other households had, so the household with the people-meter was representative of all other households in the universe, at least as far as availability and access to channels went.


But now, with households free to choose and reject individual channels, it is not necessary that a household which has a peoplemeter will have similar channels as other households in the universe.


In fact there could be thousands of permutations and combinations of households with different sets of channels. So how would one judge whether the peoplemeters households are truly representative of the universe.


Now with due respect to the existing people-meter technology for capturing TV viewership data, can we ask ourselves how geared up are we to take on the future challenges of capturing TV viewership data in light of the above mentioned technological advancement.


The story does not end here. The real challenge is still to be explained by us. TV as a medium was largely an ‘at-home’ medium. However, TV as a medium is now becoming highly portable. How are we going to measure the audience on portable TV?


For example, most of the working population may watch IPL out-of-home such as at office, at a common friend’s place, on their laptop, etc. On weekends they would also go to restaurants and pubs to watch the game in a social gathering and cheer for their favourite team. Their receptivity to the programme may be higher but it does not get covered as the people-meter is not capturing out-of-home viewership.


This may influence viewership of popular sports seasons and they may get under-reported as compared to actual viewing figures.


According to Horrell (2008), IPTV (Internet Protocol Television) could offer some encouraging answers to audience measurement such as real-time audience measurement through return-path-data technology from TV owning households itself. Moreover, it allows broadcasters to divide audiences based on socio-economic class, geographies and basic demographics. This would enable advertisers and media planners to showcase lifestyle, socio-economic class and geographic location based Ad streaming. One can avail the technology now as this is reality.


One of the sophisticated devices according to Pellegrini, Pasquale and Purdye, Ken (2004), for measuring out-of-home TV viewership is Arbitron’s Portable People-Meter for capturing passive viewership data based on encoded audio signals, embedded in the TV programmes being watched. This pager-like device would have to be carried by selected individuals to enable this device to capture the encoded audio signals. However, it could prove to be misleading if the audio signals are being captured in the vicinity but the individual is actually not watching it. For example, in a coffee shop, the PPM may record TV viewership for an individual but the selected individual is engrossed elsewhere.


What does all this mean for advertisers? How will they optimize their media investment?


In totality, TV as a medium which even today is highly fragmented with 700+ channels will have very tough and challenging days ahead as far as measurement goes.


‘The Future is now, but Measurement is Yesterday’


Extracted with permission from

The Advertising Mess

By Bipin Mundhwa and Amit Nevrekar

Published by Sci-MO

Price Rs 249, 139 pages


Post a Comment 

Comments are closed.