Banner Ad Audiences in Latin America

(photo credit: R. Soong)

In the beginning days of the Internet, one of the business models for commercial websites is based upon paid advertising in the manner of traditional media such as television, radio and print.  Given that most users still have pithy bandwidths, web advertising could not yet be based upon audio or motion video signals.  The predominant form of advertising therefore took the form of banner ads coming in standard sizes.  As with any creative work, there are good and there are bad banners, but on the whole they seem to function reasonably well.

In time, though, it is clear that there has been an inexorable decline in the click-through rates for these banner ads.  From the 1-2% in the beginning days, the current click-through may be considerably lower than even 0.1%.  If banner ads were billed on the basis of impressions (that is, the number of pages served in which the banner ad appears), then they have certainly become much less efficient than before.  But if banner ads were billed on the basis of click-throughs, then it would seem that the decline does not matter because all one has to do is to serve more and more of them.  In the long run, this is not a sustainable development, since the number of pages served and the number of users cannot be growing indefinitely.

We will now cite some data from the 1999-2000 TGI Latina study.  This is a large consumer survey conducted in seven Latin American countries: Argentina, Brazil, Chile (Santiago), Colombia, Mexico, Peru (Lima) and Venezuela (Caracas).  Out of 46,244 respondents between the ages of 12 and 64 years old, there were 8,284 Internet users.  Among these Internet users, 45% responded that they had clicked on one or more ads within the last 30 days.  From the other point of view, more than half (55%) of Internet users ignore banner ads.

The incidence of ad clickers is different along a number of dimensions.  In a previous note (Ways of Internet Access in Latin America), we noted that Internet users are qualitatively different according to the way by which they access the Internet.  The following chart shows the incidence of ad clickers by the different ways of accessing the Internet.  It would make sense that the younger school users are less interested in advertising.  It also makes sense that people who are in the 'elsewhere' category would not spend too much of their limited and possibly expensive access time on clicking ads.

(source:  TGI Latina)

The next chart shows the incidence of ad clicking according to the amount of Internet usage time in the past 30 days.  Ad clicking increases as a function of time spent using the Internet.  These data must be carefully interpreted, though.  On one hand, the heavy users view more pages, and this should definitely increase the probability of clicking on some ad at some time.  That is to say, this measure (namely, clicking an ad within the past 30 days) is a cumulative measure that should increase with the number of exposures.  On the other hand, what these data do not indicate the rate of ad clicking (say, the number of clicks per thousand impressions).  For all we know, the heavy users may have significantly lower rates of ad clicking on that basis because they have been over-exposed to and conditioned to ignore banner ads.

(source:  TGI Latina)

Here is a quick test for you: Did you notice that there was a banner ad on the top right hand corner of this page?  What was it?  Did you click on it?  If you did, you would have realized that it didn't lead to where it should ...

(posted by Roland Soong on 9/19/00)

(Return to Zona Latina's Home Page)