Italy implements a passive audience measurement with Ipsos

http://www.offremedia.com/voir-article/litalie-met-en-place-une-mesure-passive-de-laudience-radio-avec-ipsos/newsletter_id=164300/

http://fr.slideshare.net/TamaraSilina/my-concern-about-media-audience-measurements-tamara-silina-feb-17th-2013

My concern about media audience measurements Tamara Silina, February 2013

Audience measurement : What it will confront

The resistance of major media who do not want change or as late as possible the measurement systems which favor them for the audience today. The resistance of some institutes in place: A presently available device of audience measurement is a very important part of sales of some institutes. Their leaders wish to continue this activity, and for this, avoid that it evolves in a way they cannot control or would be less profitable. Organizations and professional group’s misfits: New solutions may not be supported by professional groups as they are currently organized by media type and by country. The new structure is likely to form around the new reference unit: The target advertising. The incompetence: A small part of responsibility has a global vision of the situation and real needs. Some do not have the skills to diagnose and pertinent others, faced with ambiguity of the situation, prefer the status quo "pending to see more clearly." From denial: some policymakers who have the information to understand the strategic situation can not admit the threats they involve or wish to hide them, as they did before the decision makers in other sectors (telecom, micro computer disk ...). The stereotype of consensus: it is a phase often emitted by professionals sectors who would shudder in other environments: "The current system is certainly very imperfect, but it is consensus." For some actors, the need for a common rule of a flat consensus is more important than the concerns of the right strategy, quality and efficiency. Preference for ambiguity on the part of some actors: some space buyers and sellers prefer that the quality of their services could not be evaluated rigorously and easily because they are aware of the limits of their effectiveness and, for some of them, because their practices are not for the benefit of their clients. The weight of politics, sometimes: Some institutes in some countries have provided in the past, the data which had been skewed for political reasons. The attempt to impose the watermarking: when the need for change will be binding on institutions, they will highlight the systems which benefit them in their power relations with the media.

How to start for reliable audience measurement ?

These new devices do not easily be born from existing organizations, for reasons already noted. They probably will be born in small professional environments (business clubs, sectoral committees, clubs ...) or in new countries where these structures are still in training and where the issues are too low for large institutions, or to initiative of trans-national media. Once they begin to operate, even for small applications, they may trigger an awareness and questioning will be generalized and can be brutal.

What is ideally my audience measurement system?

My vision of the ideal concept today for my radio, covers an enlarged market to my real territories (linguistics, technical, geographical ...). This market will be segmented by target advertising and editorial. It would include a market mechanism for (purchase and sale), powered by an integrated ... Data cross media audience, reliable, targeted, continuous, available in real time, including audience share of all stations likely to compete with me (so no use of watermarking). Sold data space and the advertising coverage, they also cross-media, available in real-time and instantly integrated to the data audience. Structured data and detailed information on the programs broadcast. The tools to assist decision making for both buyers space (optimization plans and impact assessment) and sellers (optimizing rates). It would be true of all technical checks by professionals such as Veritas or large international audit firms.

What say the Professionals for audience measurement?

The system which will replace those currently in place is already quite clearly outlined by the Coalition for Innovative Media Measurement. It should ideally be measured by the same source (single source) and real-time television audience and all radio channels, Internet and social networks, the press and the outdoors advertising. It will be based on data from the advertising, also available in real time. It should be exploitable across applications which support decision-making of a new generation, especially to automatically update diffusion plans but also for the media, updating the pricing space depending on their audience and demand. It will provide easily articulate on tools for measuring advertising effectiveness.

Audience measurement : What the theory says?

"Policy makers prefer alternatives with less uncertainty. This is information that reduces uncertainty. So buyers opt easier for offers supported by the latest information, most relevant, most exhaustive, most reliable, most continuous and in larger quantities. C. Shannon (1948) "A Mathematical Theory of Decisions and Communications." Conversely, buyers of advertising space in traditional media under invest because the offer is addressed to them is not supported by information whose relevance is limited, incoherent between types of media that is not new, which is discontinuous and in limited quantities. Audience measurement systems of these media and methods of buying and selling space that accompany them are handicaps for the economic development of the media and they reduce their resistance to new competition. It would be necessary to change that.

Small parenthesis about watermarking

The use of this technology obliges each station or channel that wants to measure its audience to include a code in the audio signal and therefore to pay. It is strangely almost never mentioned that the recognition of watermarking signals for radio functions poorly in noisy environments (such as in cars is one of the contexts in which the hearing of terrestrial radio is still virtually not challenged by Internet). Tests conducted on these systems have noted very low rate recognition. The television people meters have an average error of 10%. And yet, in this case, the noise can not affect recognition. The individual passive people meters tested in 2009 by Rajar had a recognition rate of about 50% Large institutions heavily involved in the audiovisual media audience measurement have been converted to the virtues of watermarking, even those who had long been criticized for its lack of reliability. The individual passive audience measurement can be performed without watermarking and with fewer errors. But that choice could lose competitive advantages to the major institutions.

Audience measurement : Other systems using mobile phones without using the watermarking

Other systems also use ordinary mobile phones and tablets panelists without using the watermarking. Their recognition technologies using channels and stations in ambient sound. These systems measure all channels and all radio stations without limitation. They measure audience share, they are in real time. Their technologies do not have hte handicap of the watermarking in noisy and parasitized environments. Some are integrated into larger systems with monitoring tools advertising and media planning, which brings more than other required system.

Arbitron PPM - MediaWatch Gfk - Ipsos Phone

Arbitron PPM This is one that has been the most concession contracts and partnerships with major institutions who have paid for not using it (like Gillette bought the patents of products that prevent the beard to grow to be not broadcast). It can provide data in near-real time by a mobile Internet connection. The fact that it is a box to always carry oneself is a handicap. "It is a not declarative system but binding and almost punitive but as the electronic bracelet of remand" as a panelist said. This is a closed system (hard and soft). Well protected by patents. Therefore, it evolves very slowly, more slowly than technology and telecom signal recognition. For this reason, he lost the advance it had ten years ago. MediaWatch Gfk It is a passive individual PPM and mobile integrated into a wristwatch. It has the advantage that keeps constantly on itself. It does not have audience data in real time because it is not on-line, you must connect it to a rack to download its measures. A small module can integrate the audience of the press (in declarative). It also uses the principle of watermarking, even if Gfk Telecontrol pushed his researches far enough in recognizing the signals without watermarking. Its designers have forgotten that a watch is a very personal object, which bears the image and style of its owner. They developed models very ugly and very recognizable. This double bias unfortunately handicap strongly MediaWatch when recruitment panelists and affecting the social behavior of these panelists identified as such by their relatives. It is also a closed system (hardware and software) and condemned to move slowly in a technological changing. Ipsos Phone This is an application for the recognition of watermarking signals which can be incorporated in current smart-phones and transmits what it recognized in almost real time. This is the device that would bring the most ideal if the algorithm did not have a capacity as low recognition in a noisy environment (discussion, engine, etc.) and if not drained quickly if the battery phone. This is not a closed system by the hardware as is the case for the two systems mentioned previously. It can take advantage of other phone features such as the location, the ability to measure the behavior of Internet audience on the phone. This is one of the three that is most likely to change quickly.

Audience measurement : New systems available today

New systems available today These are the Arbitron PPM (the pioneer) MediaWatch the Gfk, the Ipsos mobile phone and other less known. All can be used for both television and radio. They are mobile and personal. All use watermarking channels and stations. They can not measure the audience, even merged, of channels and stations not watermarked. They only measure very partially the competition and do not allow to measure audience share which is known to be the No. 1 indicator of analysts from programming content. Forcing the institutions to this technology is essentially justified because it obliges the stations to pay to be measured and it constitutes a barrier for competition. This choice is not justified by the search for improving the reliability of the data These technologies relate only to the extent of the audience measurment Institute does not offer an integrated solution with the audience measurement, monitoring broadcasting of advertising programs and media planning tools.

Audience measurement : The sidelining of the real experts

The sidelining of the real experts The experts who are approached by groups are often ex operational groups or other institutions. These are not researchers. These experts tend to reproduce what they know, that is to say, the old devices. However, there are many researchers who work and publish many of these themes. I found a lot of publications that focus on dysfunctions that I mentioned. They are not solicited by groups. Themselves usually take away from these markets. Researchers that I could ask are rather pessimistic about the chances of deep change initiated by organizations today. Some imagine rather a sudden break.

Audience measurement : The major institutes have no interest in that change

Audience measurement: Audience measurement : The major institutes have n...: The major institutes have no interest in that change Audience measurements are highly profitable and safe for major institutes. They are a s...

Audience measurement : The major institutes have no interest in that change

The major institutes have no interest in that change Audience measurements are highly profitable and safe for major institutes. They are a stable part of their business, while other activities are much more fluid and disrupted. A contract for audience measurement involves several years, he leaves a margin greater than 30% and is often extended. There is virtually no industry that has both a margin so high and so low risk. Measurements in the field of television audience, there is a technical barrier to entry. It is almost impossible to institute to win one of these tenders. In some cases, major media institutions participating in the capital that measure them. In many other areas, typical conflict of interest would be impossible. Large institutions do not pressure to break the status co, especially with new technologies that could reduce the budgets of audience research and open the door to new institutes competitors.

Major media pay and control the audience measurement

Major media pay and control the audience measurement Groups and committees who conduct audience measurement media are dominated by big media. These are necessary because they have a historical leadership but because they pay more and others. It is therefore difficult to oppose their views. The orientations of these media are naturally rather conservative and not very innovative because the old systems protect their dominance. The amount of the subscription media studies is used by major media such as a barrier to exclude small media. It is clear that if audience measurement were much cheaper, their leadership would weakened. They tend to favor heavy methodologies and denigrate innovations that could reduce costs. Small and new media will have no say in the market as long as they do not break the taboo that leadership or they will not leave groups.

Audience measurement : Manipulation of data

Manipulation of data More serious than the induced abuse by the aberrant tender specifications, data are sometimes manipulated, at the request of the media, policy makers, content producers or communication agencies. Of course, not at home or at home, but elsewhere. The Patrons of large groups located in large countries I have explained that they sometimes pay institutes have more favorable hearing. In another country, the engineers that programmed applications skinning measures of television audience explained to a reporter how the audience of public channels were routinely boosted by a discrete weighting. Moreover, media planners have detected a ratio always identical in the daily number of minutes of audience average chain from major brands for several weeks. All these examples are recent. They have not been possible because data controls are deficient.

Audience measurement : A almost complete absence of control

A almost complete absence of control Within modern markets such as exchanges, control systems are very stringent. They focus on the design of systems, their operation, their integrity and the absence of abuse in the use of data. Each of these controls is assigned to a separate institution and whose control is the only activity. For example the control systems engineering computing prices can be assigned to control a company or an international technical audit firm recognized. Strangely, control practices for audience measurement beyond these rules. The controls are organized as part of committees often people who participated in the selection of the device, or close to them. Technical assessments are not assigned to companies of technical control or audit firms whose job it is, but individual experts whose methods lack rigor. Their actions are not frequent.

Audience measurement: The time for access to data making it little use

Audience measurement: The time for access to data making it little use: More information is available quickly in a market, that market is more active. That's what I thought until I worked for the radio station. F...

Recruitment methods creating bias

Recruitment methods creating bias I was lucky to have me information from teams leading institutes audience measurement hearing. One point that is most disturbing the representativeness of the recruited for these measures. One factor through the most important is the rate of refusal to participate in the investigation or panel. But for audience measurement, much would be very difficult to recruit: 9 out of 10 refuse during telephone surveys, nearly 99 out of 100 for panels to people meters. This is far above the standards that would have to pretend that representative samples. Some profiles (upper socio-professional categories in the example) proved to be impossible to recruit. This means that some key targets would simply not represented. Although their presence is indicated in the methodological notes institutes. Contrary to what is said by some, no method of recovery can correct these problems. Samples are unfortunately largely composed of atypical people who can not realize the real hearings.

The time for access to data making it little use

More information is available quickly in a market, that market is more active. That's what I thought until I worked for the radio station. For us, the audience measurement are made using a telephone survey. About a thousand people every week questioned on what they heard yesterday. One might say that data every week, it's not bad, it is expected to stimulate the market. This is not the case. The data are processed once per quarter and it takes almost six weeks month to have access to intermediate data. So I have now the end of October, data from August. So I can not use the data to justify my fees with my new grid schedule. Advertisers can not make a decision and do not buy my space unless I sell out. Next year, we will not change the grid. Why not treat data every day?