MUMBAI: A watch to measure ratings from a cross section of mediums without being obtrusive! That is the latest innovation from the Swiss based technology company Telecontrol which is owned by media research firm GFK.
The product known as Mediawatch is currently testing in several countries including the US. As had been reported last year by Indiantelevision.com, Telecontrol has licensed its technology for television ratings measurement to Indian company Audience Measurement and Analytics (AMA). This will give ratings data the next day
Speaking to Indiantelevision.com recently about Mediawatch people meter inventor and the founder CEO of Telecontrol Professor Matthias Steinmann says, “We have developed the Mediawatch for passive metering. This is the next big step for us. The product is testing in several countries, including the US.”
Steinmann states that what is unqiue is that the product enables measurement of not just television and radio but also cinema, print and outdoors. He points out that wearing a wrist watch is the most unobtrusive way for someone to participate in a media survey. “MediaWatch technology offers the media research industry a completely new approach. To date, research has been media-specific, which means it specialises in measuring the use of one particular medium.”
“Now for the first time it is possible to carry out research into media consumption which is user-specific, and thereby record and analyse the complex multimedia mix to which people are exposed today.”
How it works: The watch has a multimedia signal source. The person simply wears the watch. Apart from measuring print where the user has to press a button the rest is automatic. Anyway print is the easiet to measure as people generally do not subscribe to more than two newspapers or magazines.
By collecting audio fingerprints and RF codes three times a minute Steinmann claims that Mediawatch can measure exposure to all media down to an accuracy of 20 seconds.
A recording unit records and encodes audio and radio signals three times a minute. MediaWatch is able to record data for up to four weeks before it has to be changed. If clients want, they can have docking stations installed for panel members which can be used to send the signals recorded during the day to the GfK IT centres overnight.
The microphone in the watch picks up the sound which is then compressed. “So if you are listening to a radio station then the data collected from the watch is compared with data from various radio stations. For cinema a voice match is found by going through the server’s database, which is located in Germany. The Mediawatch can also measure outdoor advertising. A person’s contact with outdoor advertising can be measured for the very first time.”
Going not smooth on UK radio for Mediawatch:
Meanwhile the UK’s radio industry body Rajar has concluded from tests conducted in November 2004 that MediaWatch failed to meet its requirements as a suitable audience research technology. Two devices have therefore been left in the race to win the contract
The radio audience measurement device has failed Rajar’s tests into the viability of audiometers. Rajar had put three such devices under scrutiny last November with the long-term aim of replacing its diaries with a more high-tech survey method.
In addition to MediaWatch the other two devices were the Arbitron Portable People Meter (PPM) and the Eurisko Media Monitor. The objective of Rajar’s Audiometer Validation Test was to determine how well the audiometers could identify up to 33 different radio services under a variety of ‘real-life’ situations. All three audiometers were subjected to the same listening conditions over the course of a weekend in November 2004.
The audiometers had to be able to identify all formats – whether music or speech – against various background noises, when played at differing volume levels and regardless of whether the wearers were stationary or in motion.
Rajar has said that the test results showed that the Arbitron PPM and the Eurisko Media Monitor produced very similar levels and patterns of listening, while the GfK/Telecontrol MediaWatch identified fewer listening occasions. The first two will now be subject to further fieldwork tests.
Rajar MD Sally de la Bedoyere said: “Rajar has selected the Arbitron PPM and the Eurisko Media Monitor meters following analysis of the test results, which showed that, despite using different techniques, both audiometers identified very similar numbers of listening sessions. Their level of correct identification was in line with that anticipated when designing the test. Additionally, it must be remembered, that Rajar’s budget for audiometer testing is a finite one.
“Following the results of the Audiometer Validation Test, GfK/Telecontrol MediaWatch is currently making adjustments to its editing software and increasing the frequency of data collection from three to six times per minute and has withdrawn from the Rajar test programme. However, Rajar would be happy to review the situation, if appropriate, in the future.” Rajar will now concentrate on designing the next round of tests, with a view to scrapping the current diary system of measuring radio audiences in 2007.
Responding to the development Steimann issued a statement saying, “We are highly astonished by the results. We cannot agree with the results of tests which have been presented in such a way that a casual observer will conclude that the watch registers less than one-third of all listening.
“Such claims are not credible, and in stark contrast to tests conducted by one of Germany’s most prestigious engineering institutes, and in contrast to our own experience over the past four years.
“In the Rajar tests, the performance of the three technologies was judged by their recognition of radio signals during ‘listening’ sessions with varying levels of interference. Telecontrol’s own audio engineer, present at the tests, judged that in the majority of cases, no listening should be registered. The MediaWatch is carefully designed to register listening under a wide range of normal circumstances. In our estimation, the Rajar tests placed undue emphasis on exposure to radio signals on the borders of inaudibility and beyond, which we would consider to be correctly described as ‘non-listening’. Yet Rajar is drawn to the conclusion that devices that register a greater number of these ‘non-listening’ events are superior technologies. This we cannot accept.
“We are also very concerned that the scenarios in which Rajar tested the MediaWatch were in many cases situations which are in direct contravention of instructions typically given to respondents. The MediaWatch is designed to register listening while being worn on the wrist. It is not designed to measure listening while being carried in a bag, and its performance should not be judged in such circumstances. In a survey, we would consider such behaviour to be non-compliance and we would eliminate such records from the database of listening on these grounds.”
Steinmann added that Telecontrol would be conducting a new series of repeatable, valid scientific tests of the new technology in the UK, with an independent, scientific engineering institute. These tests would test more frequent data capture processes in order to demonstrate that the system is sufficiently robust for consideration, and in order to establish an objective measure of the quality of the MediaWatch technology.
The Indian Situation: Steinmann adds that the technology will come into India at a future date. Right now the focus is on getting clients on board for the television service AMap. AMA CEO Tapan Pal admitted that, “The challenge in providing an alternative (to Tam) is that clients often dont want to spend a lot on research. When what they get for a certain low price is sub standard then they get more reluctant to spend. It is a vicious circle but we are confident of breaking it.
“We have given prospective clients a one-month trial. Before they subscribe it is important that they understand the value on the table. Otherwise how can we expect market support? It is going to be a long haul for us and we are ready for that.
“Hopefully in the next couple of weeks, deals will have been negotiated. We are at an advanced stage of negotiations with certain parties in terms of going ahead with the subscription. Doordarshan for one has expressed interest.”
The Road Ahead: As far as the growth roadmap in India is concerned Pal states that 2,000 peoplemeters have been installed. “We are now able to give you data for Mumbai, Delhi, Ahmedabad and Kolkata. We will also be able to provide data for 1 million plus towns in MP, UP, Punjab, Haryana and HP. Our next push will be in the South. We are aiming at a sample size of 20,000 over the next couple of years.”
Consistency in measurement of paramount importance: When asked about the changes that had been noticed in the way television audiences are measured over the years Steinmann says that it is important to have multi variant systems. “Our plug and unplug system involves about five to six different kinds of measuring that have been adapted to frequency measurement, encoding and the boost. In Switzerland we have two way coding.
“We never change the household we are in. When a household’s viewing patterns get complicated we put in more sophisticated peoplemeters for them. In India we have put the more sophisticated meters in SEC A+ homes. After all the same sample to measure a mass channel like Star Plus cannot be used to compare HBO and Star Movies.”
Talking about the health check that is place Steinmann says, “The technical support team can look at the performance of the peoplemeters from the laptop. Recently the technical team came out with a 90-page report. Using this we give AMA feedback on how their system can be improved including areas like distribution. That is what keeps me coming back to India. Research we do must reflect reality. That is our only target. However in social science you can never be 100 per cent accurate.”
Another problem in India is that in the non-addressable environment you have the situation of cable operators blocking channel signals and vice versa. When asked how aMap would be able be able to adapt in this environment Pal says, “If the channel has been blocked, it does not get recorded. However a note of this is kept when channels do analysis of viewership trends six months or one year down the line. Another situation could be that due to rains in Mumbai people stayed at home and so viewership in the day went up abnormally.
“You do not want to take aberrational situations into account when looking at viewership. This information should be eliminated when doing an analysis. On the excel title sheet the aberrations will pop up. The users have the option of keeping or removing the aberrations.”
Dwelling on the challenges he encountered when inventing the Telecontrol system Steinmann points out that in the early 1980s there were no people measurement systems. The idea was to use a box and a remote control connected to the television. “In 1984 and 1985 we introduced systems in Germany and Switzerland. Nobody had done it before. Meanwhile Nielsen and other companies also took to the idea.”