Advertisement noting

What ad noting measures, and its limitations

Selective perception, or ‘involvement’, which underlies the previous chart on product field interest, also explains some of the findings produced by advertisement noting techniques. Information is sometimes sought on the chances of an advertisement being seen if it is on a right-hand page versus a left-hand page; in colour or black and white; in a large size or a small size; at the front of the magazine versus the back of the magazine; and so on. Ad noting scores have long been offered as a way of providing some answers, though the technique has its limitations.

The method is based on showing a sample of readers a copy of the issue they have read, and asking them to go through the issue page by page and stating what they can remember having looked at previously. This can generate a spread traffic score (the percentage of readers who looked at anything on the two-page spread), a page traffic score (the percentage who looked at anything on the page) and an ad noting score (the percentage who looked at the advertisement). Averaging across all advertisements, or all within a given category, produces average ad noting scores.

The level of the ad noting scores is dependent on the form of the question asked. A question which merely asks whether or not the reader looked at the advertisement is likely to produce lower scores than a question which distinguishes between (a) just glancing at it and moving on to something else, and (b) actually reading something in the ad.

This is because most respondents can’t believe interviewers count casual screening-out of advertisements as ‘looking at’ the ads, so they don’t claim such glancing unless specifically asked about it. It makes a big difference to the scores. The JICNARS “Reader Categorisation Study” [34] has already been cited, which distinguished between “saw and read something” on the page and “saw but just glanced at” the page; combining both types of exposure yielded average page traffic scores of 90% or more.

As far as advertisements are concerned, one of the few published examples which made this distinction relates to ads which appeared in an issue of Big Farm Weekly, a publication for farmers [45].

In a traffic and noting study conducted by Gallup, informants who failed to claim to have looked at selected advertisements were asked what they thought the reason for not looking was. This established that the great majority of non-noters had in fact looked at the advertisements but had not originally claimed to have noted them because they had not felt interested in them - often because the product was not relevant. These farmers had perceived the ads in order to decide not to study them. The breakdown of the scores for one of the advertisements, for a herbicide for barley fields, was as follows:

Although the ad noting score was only 34%, 80% of informants had actually looked at the ad. Even this was not the whole story because the spread traffic score was 82%, which is a better estimate of the proportion who had had their eyes open in front of the ad.

Our understanding of ad noting scores was assisted by a validation study by the Agencies Research Consortium, a group of about 30 advertising agencies [46]. BMRB developed a technique called DEMOS (Direct Eye Movement Observation System) in which respondents waiting in a waiting room read a magazine or newspaper on a lectern, while their eye movements and the page that was opened were secretly filmed by two hidden cameras. By superimposing the film of eye movements onto the film of the opened pages it was possible to analyse exactly where on each page the eyes were directed.

Once the filming was completed the respondents were shown into another room and given a traffic and noting interview, using the same issue of the publication they had been observed reading in the waiting room. It was then possible to compare the reading claimed in this interview with the reading observed from the films.

There were some dramatic differences, largely in the direction of under-claiming in the personal interview. The page traffic and ad noting data were clearly not measuring exposure but communication of some sort. The scores were reflecting interest and involvement in the subject matter of the article or advertisement, as opposed to exposure to the page or ad (which was typically much higher).

Another reason for page traffic and ad noting scores being under-estimates of exposure concerns the age of the issue at the time the interview was carried out. If the interview is conducted too soon (e.g. a day or two after a magazine is published) the reader may not have finished reading the issue, and thus may correctly not claim to have read a page which would actually be read a little later. If the interview is conducted too late (e.g. two weeks after a weekly magazine is published) the respondent’s memory may have faded, and there may be confusion with reading of the subsequent issue. There is no perfect timing.

Indices of ad noting, by size, colour and other factors

Nevertheless, average noting scores broken down by type of advertisement can give an indication of the relative effect of size, colour and position of advertisements - provided one bears in mind that the indices are not reflecting exposure to the ads, but recall of (and thus involvement with) the ads. If the sample and variety of advertisements is large enough, the effect of other variables - especially creative execution and product field - are averaged out.

Noting scores are no longer researched regularly in the UK, but in 2004 Media Dynamics Inc compiled averages based on studies conducted in the US over a number of years by Burke, Gallup & Robinson, Starch and other research organisations [47]. The resulting noting scores were indexed with the score for an average four-colour page advertisement as 100.

The results show a logical progression by size: the larger the advertising space, the greater the recall score. Full colour outscores two-colours, which in turn outscores black & white in the larger spaces but they are level in the smaller spaces.

Additional American information comes from Starch Tested Copy data from the Roper Starch research agency during the period 1981-1990, based on measuring a range of consumer magazines [48]:

The indices show little difference on average between left-hand and right-hand ad pages. Bleed pages (ads printed to the edge of the page with no margins) outscore non-bleed, in both colour and black & white. The physical size of the page does not make a significant difference, comparing Reader’s Digest-sized pages and the larger standard sizes. Early ad pages in a magazine tend to score rather better than late pages on average, but it is heavily dependent on the specific content of each page – which reminds us that these are recall scores, not true exposure scores.

New data were published in 2005 by Medialogue, the advertising sales house of Sanoma Magazines in Belgium. Their report “Stop/watch” [49] published scores based on 2,879 advertisements tested in mass-market magazines in Belgium during 1996-2004. Ads were scored on eight measures, but the one relevant in this context is the Starch-like recognition score. Scores below are indexed, with the top item in each block indexed as 100.

In general the indices move in the same directions as the American data, though not always to the same degree. The principal difference is that Medialogue reports only a negligible difference between the first, second and final thirds of the magazine whereas Starch reports more substantial differences.

Medialogue adds an interesting new analysis: there is a modest advantage in being among relevant editorial content, such as a food product being advertised within a cookery feature. This score is about recognition of the ad; there are further advantages in the way the ad is experienced by the reader. While engrossed in a specialist section such as cookery, motoring, fashion or home interiors, the reader is already in the mood to be receptive to an ad for a product or service in the same field. (See section 15 for more on this.)

Ad clutter is not a problem in magazines

Advertisers are understandably concerned about advertisement ‘clutter’ in the media – increasingly so, as the volume of advertising messages escalates. With television and radio the linear nature of the medium means that if the length of the commercial break or the number of different commercials in it are high, it is more difficult for any one advertiser’s commercial to be noticed and attract the attention of the audience.

There is no such problem of clutter in magazines however. The targeted nature of a magazine’s readers means that most magazine ads have some degree of relevance to the issue’s audience. Indeed with most magazines the advertisements are regarded as an integral and important part of the content. It is then the readers who control the order in which ads are looked at, and the attention and time devoted to each one. Moreover advertisements are distributed more evenly throughout the medium than is the case with television and radio.

Consequently a given advertisement’s impact is not likely to be much affected by whether there are other ads nearby, ads from direct competitors, or whether a high proportion of the issue’s pages are made up of ads. This has been confirmed by the “Stop/watch” report [49] published in 2005 by Belgium’s Medialogue (described above).

There is no advantage (in terms of recalling the advertisement) in being physically distant from all other advertisements. Ads with other ads close by achieve more or less the same average score as ads that are alone. The difference between an index of 100 and 101 is not statistically significant.

Nor is it a drawback to have direct competitors advertising in the same issue of the magazine. It makes no difference whether there are no competitors or eight competitors in the issue. This is a contrast to television and radio, where it is a cardinal rule not to have direct competitors in the same commercial break. (The magazines tested were mass-market titles. With specialist magazines, readers positively want lots of competitors in the same issue. Computer magazines are a good example: many are thick with advertisements and catalogues, forming a marketplace.)

Finally, it is no disadvantage to have a high ad ratio: that is, the proportion of total pages in the issue which are advertisements. Indeed, quite the opposite. For ads in issues where 30%-40% of the pagination is advertising, the average score for an ad is slightly higher than for issues with lower proportions of advertising.

Clearly, ad clutter is not a problem with magazines.

Eyes open in front of page: the real measure of net audience to ads

Readers generally open all or almost all pages and are thus exposed to the ads as well as the editorial, which means that the publication has done its job of putting the ads in front of the eyes of the audience. The Reader Categorisation Study (discussed in section 10 above) showed that 90% of readers looked at the average page in a magazine. The limitation of the Starch and almost all other page traffic and ad noting techniques is that it is not clear to respondents that just glancing at the page counts as exposure to the page or ad. The readers think they are only to report the pages or ads that made a memorable impression at a conscious level – which depends on the attributes of the advertisement (such as creativity and likeability) and also the reader: the reader’s needs and interests, attitudes towards the brand, and experience of other communications from the brand.

The high spread and page traffic scores from the Reader Categorisation Study have been echoed by a German project [50]. Measuring eye movements, it was found that eye contact is made with approximately 90% of all advertisements.

The figure of 90% of readers looking at the average page is simply a measure of net audience, of course. The gross audience takes account of repeat reading and is considerably higher, as section 11 described.