Attention Levels to Advertising on Different Media

(photo credit: Roland Soong)

The quantitative science of advertising is premised upon the existence of accurate databases to which complex mathematical and statistical tools can be applied.  Since big advertising campaigns involve multimedia approaches, accurate data are required from various media sources.  Since different media have different characteristics, there is in fact no omnibus measurement system that can simultaneously provide data on the different media.

Here are some typical measurement systems:

• Television  Once upon a time, in a much simpler environment, people were asked to keep diaries of their television viewing.  The obvious shortcoming is that the diary-keeping interferes with normal viewing behavior and that the recording accuracy and completeness deteriorates over time.  Another system is based upon (unaided and/or aided) recall of yesterday viewing, but the shortcoming is that it is subject to memory failures, especially for programs and stations that are not in top-of-the-mind awareness.  The performances of the above systems have worsened in a complex environment when people have over hundreds of television channels to choose from Today, the most common system is one that is based upon a panel of television households which have been equipped with electronic meters to capture household tuning and personal viewing, from which television ratings for television channels and programs can be delivered on a near-real-time basis (see James G. Webster, Patricia F. Phalen and Lawrence W. Lichty (2000).  Ratings Analysis: The Theory and Practice of Audience Research.  Lawrence Erlbaum Associates: Mahwah, NJ).
   

• Radio  The dominant method would still appear to be to ask people to keep diaries of their radio listening.  Since much of  radio listening takes place outside of the home, these cooperators are usually ask to carry their diaries everywhere they go.  Obviously, this may disrupt the normal listening behavior.  In practice, some of the data are reconstructed by the cooperators via recall.  The audience information is delivered as ratings for individual stations by time of day.  The radio audience information is usually broken down by demographics such as age, sex and socio-economic level (see, for example, Radio Stations & Format Genres).
    

• Newspapers  A common, but by no means universal, technique is based upon first showing the logos of the newspapers and asking a time-related filter question, "Have you read or looked into any of these newspapers in the last X days/weeks/months?"  For those newspapers that passed the filter, the readership question is based upon "When was the last time before today that you read or looked into an issue of this newspaper?"  The newspaper audience is defined as those who read yesterday.
   

• Magazines  A common, but by no means universal, technique is based upon first showing the logos of the magazines and asking a time-related filter question, "Have you read or looked into any of these magazines in the last X days/weeks/months?"  For those magazines that passed the filter, the readership question is based upon "Did you happen to read or look into any of these publication in the past week/two weeks/month, not including today?"  For a weekly magazine, the audience is defined as those who read within the last week; for a monthly magazine, the audience is defined as those who read within the last month; etc.
   

• Internet  Since there are millions and millions of websites on the World Wide Web, there is no hope of using any method based upon recall or concurrent diary keeping.  Instead, the most common systems are based upon software 'meters' that record user activities for a large panel of cooperating households.

Unfortunately, this is not the whole story, even if these audience statistics are the numbers that are quoted most often in the popular press.  Knowing the number of people who read a newspaper or magazine, listened to a radio station, watched a television program or viewed a web page is still insufficient because one should consider what level of attention this audience is paying to the commercials embedded within the media content.

We will now cite some survey data from the TGI Mexico survey.  This is a consumer survey of 11,400 persons between the ages of 12 and 64 years old conducted by Moctezuma y Asociados in Mexico during 1999-2000.  Within this survey, we asked to respondents to rate their attention to advertisements in various media on a five-scale, with 5 meaning always.  From this survey, here are the percentages of people who give the highest rating of 5 to each medium.

Attention Levels to Different Media by Persons 12-64

These attention levels are in fact different by sub-population.  It would seem patently unfair to be basing this comparison on the total population, when not all elements of the population use all of these media.  In particular, we should wonder what would happen if this table were restricted to the population of Internet users (defined to those who have used the Internet in the past 30 days).  When we compare the group of Internet users against the general population, we see that they are tautologically more attentive to the Internet, but also that they are less attentive to broadcast media and more attentive to print media.

Attention Levels to Different Media by Internet Users (past 30 days)

In a previous note, we had pointed out the differences between people who use Internet for entertainment versus for informational purposes.  Given their different goals, it is conceivable that they might have different attitudes towards advertising.  The following two tables show the attentional levels for these two groups of Internet users.

Attention Levels to Different Media by Entertainment-seeking Internet Users (past 30 days)

Attention Levels to Different Media by Information-seeking Internet Users (past 30 days)

In the case of the entertainment-seeking Internet users, we actually see that the Internet is the medium that commands the most attention from them! 

WWW LINKS

• Zona Latina's Internet research articles

(posted by Roland Soong on 10/29/2000)

(Return to Zona Latina's Home Page)