Commentary

TV Ratings Debate Focuses On The Wrong Issues

In the middle of the Nielsen debate, isn't it high time everyone took a step back and looked at some of the larger issues that have been raised?

The industry is currently such that only some media can be measured precisely, and only some advertisers can target their audiences specifically. Therefore, why should clients pay for media when they know that 50 percent of their spend works, but they don't know which 50 percent? What other industry worth $50 billion would survive on such thin ratification of such a huge expenditure?

The concept of using a representative sample is the basis of the market researcher's art. However, the level of accuracy is poor when predicting a two-horse political race, let alone when trying to map patterns over several hundred variables such as TV channels. One has to have sympathy for Nielsen; however, evidence shows that the sample-based model is a hugely flawed approach when researching media.

How do I know this? Quite simply, in addition to my experience at Ogilvy & Mather and CIA Interactive (now Mediaedge:cia), two years ago I helped to develop a technology called Adserver* that measures not only the number of commercials viewed, but also the amount of time they are watched on a viewer-by-viewer basis. Adserver* currently operates on online channels such as extreme sports channel High.tv (http://high.tv) for clients such as Nissan, Salomon, AOL, and SonyEricsson and on Global Sport TV (http://globalsport.tv) for clients such as Nike.

advertisement

advertisement

Nielsen's NetRatings figures for High.tv, using a sample panel, are half the achieved rate indicated by the database (incidentally, a 30-second ad boasts a 97% view-through rate, and a 2-minute ad a 67% view-through rate--something Nielsen can't tell you). I suspect that inaccuracies like this will also creep into the broadcast model.

Let's put Nielsen measurement inaccuracies aside and focus on a universal and yet largely ignored issue in the media business--the tendency to exaggerate actual delivery in favor of reach.

The whole model is skewed toward reach, not accuracy. The whole media business is based on delivering volume, not targeting customers. A sponsor will quote the reach of a channel a program appears on rather than the exact viewing figures. Thus, most figures are exaggerated. Equally, we all know that people channel surf or make a cup of coffee during the commercial breaks. Why else has the annoying practice of increasing the volume of commercials become prevalent? Or, indeed, the use of sponsor slots over traditional commercials?

This shotgun approach, though, has had its day. Clients are more than willing to entertain ever more targeted media, often with a very low reach--it is their buying agencies that are unwilling to change. But the writing is on the wall. Google has created a closed advertising community that has locked out media agencies, and this is going to happen more and more.

Ironically, sampling issues such as race or gender are irrelevant in the Google model. Similarly, clients using our Adserver* can already book their commercials directly for playout on online channels. They can adjust their campaigns, alter frequency, reach, and payment levels, and monitor their conversion levels in real time. Soon, this will be available for commercials' delivery on cable stations.

Nielsen should be commended for trying to improve their sampling, but no matter how good the sample, it will always be skewed against some interest group. In the meantime, the client is likely to want to get 100% of their advertising spend working, and will turn to media that are truly accountable.

Iolo Jones is the CEO of TV on the Internet Company, Narrowstep Inc. (www.narrowstep.com). He previously worked for Ogilvy & Mather and CIA Interactive (now mediaedge:cia). Iolo can be reached at ijones@narrowstep.com.

Next story loading loading..