Beyond the OTS: measuring the quality of media exposure.
Media owners, advertisers and their agents in just about every country around the world invest large amounts of money in measuring the size and composition of audiences exposed to a wide range of media vehicles. These media research studies include peoplemeter systems for measuring television audiences, readership surveys of newspapers and magazines and radio listening diary studies. The biggest use of all of these surveys is in the planning, buying and selling of advertising time and space. At the most basic level they provide a trading currency for the advertising industry.
In general, these media research studies attempt to provide an accurate measure of the numbers and types of people who have an 'opportunity to see' (OTS) an advertisement placed by an advertiser. Typically that would mean someone present in the room at around the time a particular TV commercial was shown or someone who in the past read or looked at any part of an average issue of a publication in which a particular press ad will be placed.
This article concentrates on research which goes beyond these 'opportunities to see' and attempts to measure something more closely linked to the likely effectiveness of the advertisement. There are two basic questions to answer here:
(1) How do we determine a better measure of who has actually looked at a specific advertisement placed within a medium?
(2) How do we determine which people are in an appropriate frame of mind to respond to the advertisement?
Together these areas are generally referred to as the quality of media exposure.
The years of enlightenment (the 1950s-80s)
The industry has long recognised the inadequacy of the primary 'OTS' media audience measures as gauges of the likely impact of advertising. One only has to look at compendiums such as the old Newsweek Media Research Index (last updated in 1991) and into the WARC.com archives to see that people have been conducting studies to investigate these questions since the 1950s. The following selection gives you a flavour, rather than a comprehensive account, of the extensive body of work that has been built up over the last 50 years.
In 1956 the Advertising Research Foundation (ARF) published A Study of Printed Advertising Rating Methods, a milestone study which established that pass-along and out-of-home reading had 85%-90% of the value of primary or in-home reading.
As early as 1962 a series of studies by the Home Testing Institute in the US showed that attention to programmes and commercial recall were greater among those viewers who considered a programme one of their favourites. Then in 1965, Ogilvy, Benson & Mather showed there was a difference in recall levels between advertising in different dayparts on television in their Experimental Study of the Relative Effectiveness of Three Television Dayparts.
By 1968 W.R. Simmons published the findings of his Pilot Study on Immediate Recall of Television Advertising to show that advertising recall levels were far higher among viewers paying full attention to a programme. At the same time, Thomas E. Ryan was completing his A Study of Reading Environment and Advertising Memorability, which showed that there was a relationship between 'degrees of involvement' of magazine readers and the percentage who remembered specific ads.
During the 1970s, Sonia Yuspeh's On-air: Are we Testing the Message or the Medium? advanced the debate by showing that not only was commercial performance affected by programme environment, but also that performance varied significantly between different shows of the same programme type. While Ray & Webb (1979), using laboratory tests, found that first position in a string of commercials was associated with higher attention and recall, and that as the amount of clutter increased there was a consistent decrease in effectiveness scores.
Clutter was also the issue that Consumer Dynamics Corp. grappled with in their study for Esquire on A Study on the Effects of 'Clutter' in Magazine Advertising in 1980. This showed that each of four test ads on average received 26% lower claimed recall (and 22% lower unaided proven recall) in a cluttered version of the magazine as opposed to the regular version.
During the rest of the 1980s people continued to grapple with the issue of how media environment impacted on advertising effectiveness. Elliot Young (1983) used eye tracking studies across eight magazines to show that a reader was likely to 'be involved with' 58% of ads in a publication average, with involvement levels ranging from 78% to 2% in the eight magazines.
Clancy, Shulman Associates published their Program Impact & Program Appeal: Qualitative Ratings & Commercial Effectiveness in 1984. They confirmed, by exploring the relationship between measures of 'Program Appeal' and 'Program Impact' and four measures of commercial effectiveness (brand name recall, message recall, message credibility and intention to purchase), that viewers' involvement with a programme does affect their response to commercials.
Then, in 1986, The Impact of Editorial Environment on Brand Acceptance -- A 13 Magazine Analysis Study from Lieberman Research compared the effect of editorial environment on consumer reactions to magazine advertisements.
Therefore by the end of the 1980s there was a fairly broad understanding and acceptance of the key influences on quality of media. These could basically be categorised under the two headings of 'Attention Effects' and 'Clutter Effects', and they were issues common to virtually all media. 'Attention Effects' revolved around the level of attention consumers devote to the media opportunity a commercial is placed within and the degree of their involvement with the media. It had always been clearly recognised that there were big differences in commercial effectiveness between different media which were largely recognised in their relative price. However, perhaps more importantly, we had established through myriad studies and tests that there were also big differences in effectiveness between opportunities within the same media due to attention and involvement levels.
'Clutter Effects' were those where the impact or effectiveness of advertising was influenced by the amount of other commercial activity taking place around it within a given media opportunity.
From enlightenment to actionability (the 1990s)
During the 1990s the investigations into media quality have continued. Some studies have reiterated work that had gone before, but reflect the modern media context with its explosion of media opportunities and increasingly demanding consumers. Others have focused on a weakness with nearly all of the earlier work, which was that, whilst it had delivered useful learning about the way that the media environment influenced advertising effectiveness, it provided little actionability for media planners on a day-to-day basis.
The rest of this article will concentrate on a range of studies that were carried out during the last decade of the second millennium. In part these studies have sought to investigate further quality of media and verify whether the earlier findings still apply in the more complex modern media environment. In part, and perhaps more significantly, they have sought to advance things by providing a basis for planning and negotiating media opportunities by applying 'Quality of Media' measures.
At Carat we instigated a research programme that was intended to get to the bottom of the key media quality issues and then provide actionable insight and tools for our planners in the early 1990s. The programme is still ongoing but media quality is already a primary planning tool across the group. We carried Out two primary pieces of work into the quality of TV viewing, which together formed our Foretel Study. The first was the Foretel Advertising Recall Study, the second our Foretel Attention Study.
The Carat Foretel Advertising Recall Study, 1993
Through Broadcast Research, we recruited five samples of 200 adults who were representative of ITV/Channel 4 viewers in each of five dayparts (making 1,000 respondents in total). They were given sealed envelopes containing a questionnaire on ad recall and a short diary of TV viewing, and were told to open these at a specific time on a specific day.
The first exercise they had to do on opening the envelope was to write down all of the commercials they could remember seeing in the previous two to three hours. To jog their memory a list of product fields was given. Their next task was to fill in a viewing diary indicating which ITV/Channel 4 programmes they had watched, whether they enjoyed them, whether they especially chose to watch and their attention levels by quarter hour.
This technique allowed us to measure spontaneous recall of advertising within around two hours of transmission, and to tie this in with the quality of people viewing at the time of exposure. The database that resulted from this study consists of over 12,000 observations across over 200 commercial transmissions. By looking at averages over such a large database, we are confident that biases due to factors such as creative execution are ironed out.
An example of the data is the centre break of the popular Australian soap, Home and Away (see Table 1). The first ad in the break, for Cadbury's Boost, was correctly recalled by 18% of the adults watching at that time. This was the highest score of the break. The last ad in the break had the second highest recall score (16%), and the ads in the middle seem at first glance to be recalled in line with spot length.
We can now use these data to examine the buyer's prejudices. Is 'first in break' really the best position?
Position in break
For the Home and Away break the answer seemed to be yes. The average recall for the break was 13% but the first ad was recalled by 18% of Home and Away viewers. That represents an improvement of 38%. The last position is 23% better than average and middle positions are 15% worse.
These results could not be regarded as conclusive. The findings for this break may be entirely due to creative impact or the weight of previous media exposure for the individual commercials in question. We needed to look at the full database to draw reliable conclusions. When we did this we saw a very different picture (see Figure 1). On average, the first ad in the break was recalled by 9% fewer people. Far from being the best position it could possibly be the worst. The buyer's prejudice is wrong.
It also seemed to be wrong about centre breaks. Our research showed that the average commercial was recalled by 16% of those exposed to it, irrespective of whether it appeared in a centre break or an end break. Therefore peoplemeter ratings are an adequate tool for deciding between centre breaks and end breaks.
Time of day and size of break
What peoplemeters cannot tell us, however, is that quality of viewing varies dramatically by time of day. Far fewer viewers recalled ads shown in the late afternoon than at other times of day. Against the buyer's 'gut feel', the late evening segment performed best. This is probably because people watching after 10 p.m. have made a special effort to do so rather than go to bed.
The buyer is not always wrong. We found that recall of ads in very short breaks was higher than longer breaks with more ads. It is interesting to note, however, that being one of six ads in a break does not give any better chance of recall than being one of 12 (see Figure 2).
The key area under investigation with this research project was how programme environment affects ad recall. As part of the viewing diary that respondents were given, we asked people to rate each programme they had viewed on a 5-point scale, where 5 meant they enjoyed it very much and 1 meant they did not enjoy it at all. The results showed that, on average, people who enjoyed a programme recalled 20% more commercials within or after it.
Enjoyment was not our only measure of viewing quality, and, in fact, we found that people's reasons for watching the programme were an even greater influence. People who specially chose to watch a programme recalled, on average, 18.3% of the commercials, whereas people who watched for other reasons recalled only 13.5%. The increase is 35%.
The biggest influence on recall was the level of attention paid to the programme. The attention scale we used in this study asked respondents to describe their level of attention for each quarter hour of their viewing, using a 4-point scale:
* Full attention. Looking and listening to all or almost all that happened.
* Average attention. Paying attention to most of what happened, but also doing other things.
* Little attention. Looking at the TV from time to time, but mostly doing other things.
* No attention. Although the set was on.
Recall of commercials varied between each of these attention levels, with fully attentive viewers recalling, on average, 19% of the ads and those paying no attention recalling fewer than 8%. It is a testament to the intrusiveness of television that these people recalled any ads at all. But the fact remains that people who paid full attention to the programme did recall 25% more commercials than those paying no attention. Broadly speaking we can say that people paying full or average attention recall 70% more commercials than those paying little or no attention. Programme environment therefore has a massive influence on ad recall.
The Carat Foretel Attention Study Programme
If programme environment is so vital for effective advertising, the next question is which are the right programmes to be in? Carat's programme of Foretel Attention Studies has been designed to answer this question.
Carat has been conducting Foretel research in the UK since 1992. To date we have interviewed well over 30,000 adults about their quality of television viewing. The Foretel technique has seen several variations as we have focused on various issues of interest, over sampling specific target audiences, channels or programmes, or adding questions of topical interest. The typical methodology is as follows. For each sweep BMRB recontacts around 4,000 adults who have previously returned a full TGI questionnaire and identify every programme they have watched in the previous 24 hours. Before the telephone interview we mail each potential respondent within 'satellite homes' a copy of the Radio Times for their region and use this as a prompt to ensure that we pick up all of the programmes they watch. Other respondents are asked to get a copy of the newspaper or programme guide to use as a prompt. Those who cannot find one are prompted with programme names. For each programme the interview goes on to establish reasons f or watching, enjoyment, other activities whilst watching and attention levels.
The Foretel attention research has given us so much information that any attempt to describe 'the results' can only scratch the surface. Carat would obviously like to keep much of the information confidential so that it can be used to improve clients' campaigns without the knowledge of their competitors. What follows, therefore, is a brief illustration of the richness of the database that we have created.
Firstly, let us move straight to the heart of the research. How much attention do people pay to the TV programmes they watch in the UK? There is wide variation by programme, channel, daypart and target audience. The average results for ITV show that 50% of adults pay 'full attention' to the programmes they watch during peak time on weekdays. At other times of day attention can fall as low as 15%. For housewives with children the results are even more extreme with up to 97% of these busy mothers paying less than full attention whilst they watch TV. Of course, it is not reasonable to expect everyone to pay full attention all of the time, but what we certainly want to minimise when we place advertisements for clients is the chance of paying to reach an audience paying little or no attention. For housewives with children this ranges from 19% to 90% as a consequence of the time of day that an ad is transmitted.
Foretel can also be used to give a health check on the various TV channels. Our research showed a slight decline in the attention paid to ITV over the mid-1990s, although they reclaimed some lost ground in the last period (see Table 2). Channel 4 seemed to have built attention levels gradually over the years of our study, whilst the satellite 'repeats' channel UK Gold fared even better until a poor result in 1998. The breakfast time franchise changed from TV-AM in 1992 to GMTV in 1993 and showed an improvement in attention as a result, which it then regained and improved after a hiatus during 1995-6.
Individual programmes can also be handled in the same way, which tell us that whilst Britain's favourite soap Coronation Street has remained stable, others such as Emmerdale and Home and Away have declined in attention.
Foretel research indicates very strongly that, although the time of day and channel has a major influence on quality of viewing, the programme itself is of the utmost importance. For example, back in 1992 ITV scheduled a one-hour drama at 9 p.m. on a Tuesday evening called Boon. In 1993 Boon was axed and replaced with Soldier Soldier, which continued to occupy the slot in 1994. This programming change was a good move by ITV as the audience increased by 2-3 million viewers in the first year, and importantly the quality of viewing improved, resulting in 3.3 million additional attentive viewers. This represented a real bonus for advertisers and also for ITV: the better quality programme continued to improve in both audience size and viewing quality as the Soldier Soldier audience matured (see Table 3).
Why did viewers pay attention to Soldier Soldier? The answer lay in the fact that much of its audience especially chose to watch and then thoroughly enjoyed what they saw. In Foretel analysis we refer to this kind of programme viewing as a 'magic moment'.
Other programmes are not quite so special but nevertheless become a 'regular date' in the evening. Then there are the 'chance meetings', those programmes that we watch when there is nothing better on or because someone else wants to watch them. Bottom of the list is 'wallpaper TV', when people are engrossed in some other activity but still count as viewers because they are present in a room with a set switched on. We used Foretel to define these viewing types as the Foretel Viewer Involvement Profile (VIP):
* Magic moment. Specially chose to watch the programme and rated it 5 out of 5 for enjoyment
* Regular date. Specially chose to watch and rated 4 or less for enjoyment
* Chance meeting. Didn't specially choose -- but excluding 'Wallpaper' viewing
* Wallpaper TV. The set was on but I wasn't really watching.
The value of the VIP segmentation can be illustrated when comparing the quality of viewing to two game shows, Lose a Million and Fifteen to One (see Table 4).
More than half of Fifteen to One's audience regarded it as a 'magic moment', but by contrast only 5% of the Lose a Million audience fall into this category. The reason for this is that, although Lose a Million had a much larger audience than Fifteen to One, the majority of viewers were watching because there was 'nothing better on'. They were, in fact, waiting to watch the next programme in the schedule, Coronation Street.
By applying the VIP segmentation to all programmes within a channel it is possible to compare the quality of viewing to channels as a whole. This shows us that the main terrestrial channels, ITV and Channel 4, generally have the highest quality of viewing, with the majority of the audience especially choosing to watch. Overall, though, only 20% of TV viewing falls into the category of 'magic moment' viewing.
The themed satellite channels have different characteristics. Sports channels provide a high content of 'magic moments', whereas general entertainment and movie channels tend to be 'chance meetings'. News channels are either a 'regular date' or 'wallpaper TV', as (unsurprisingly) are children's channels for an adult audience.
For MTV and the other music video channels, what may be alarming is that even younger viewers (aged 15-24) regard it as 'wallpaper'. Foretel reveals that younger people are, in general, less involved with TV. It seems that the older you are, the more likely you are to regard TV as a source of 'magic moments'. One reason for this may be that younger people do a lot more with their lives than watch TV, but another is probably the conservatism of UK programme schedulers.
To use this information in media planning, we can analyse the 'magic moments' for any kind of target group and then buy into or avoid programming accordingly. Table 5 shows some examples for a very important group for advertising: new car buyers.
The main purpose of selecting 'magic moments' for a TV schedule is of course to maximise attention as well as involvement. There is a very strong correlation between VIP and full attention (see Table 6).
At Carat we developed systems and technology to optimise directly the attention paid to a schedule of TV commercials using the Foretel research, firstly through daypart and channel mix, secondly by selecting the most appropriate programmes and programme genres.
Foretel did not replace BARB as a TV buying tool, but the combination of BARB and Foretel provided a very reliable basis for planning and buying television schedules with the objective of maximising the effects of advertising. Its findings have also been supported by similar studies conducted by our sister companies in Germany and the US, which are now being used and applied to the planning process in those countries as well.
More recently, we have analysed BARB viewing patterns against Foretel information and identified that we could accurately predict VIPs for programmes based on the patterns of switching in and around a given programme. Consequently, we have now built proprietary models which provide Quality Rating Points (QRPs) for all programmes and which are built into our planning systems.
Other quality of TV viewing studies
Carat has not been alone in continuing to try to move forward our understanding of quality of TV viewing and in managing the implications from this work.
In 1995, Lloyd & Clancy published their seminal study 'CPMs versus CPMIs: implications for media planning'. In it, CPMs (cost per thousands) and CPMIs (cost per thousands involved) were compared in regard to whether media planners and buyers would make the same or different media buys if their decisions were based on (traditional) CPMs as opposed to (modern) CPMIs. They found that TV programme involvement appears positively, causally linked to advertising effectiveness.
In 1993, The Billett Consultancy's Share of Break Study identified three key media factors which determined TV advertising effectiveness:
(1) Length of commercial
(2) Number of commercials in break
(3) Position in break (first being best, but with far more marked differences in breaks with more ads).
The finding on position in break contradicts some of the findings of Carat's Foretel studies. However, further Billetts analysis also showed that commercials in the middle of short breaks performed significantly better than commercials in the middle of long breaks which turned out to be the worst of all positions.
Billetts then built on this finding with its Commercial Break Ecology study in 1998, which looked at consumers' behaviour during TV advertising breaks and evaluated its impact on recall. It developed measures of involvement and attention based on the classification structure pioneered by Carat in our Foretell TV Advertising Recall Study (above).
Billetts identified that only 32% of viewers who claimed to see a commercial break did in fact see it all, with the remainder displaying a variety of non-attentive behaviours. Even after allowing for the fact that respondents accurately complying with the peoplemeter task would be excluded from a measure of viewing, and a meter tuned to a short persistence period would control viewing measures for channel flickers, only 59% would have seen all of an average commercial break (see Table 7).
Billetts went on to confirm Carat's findings that programme attention levels had a major effect on advertising recall.
Then, in 1997, Xinshu Zhao published a study of the impact of clutter in the Journal of Advertising Research, entitled 'Clutter and serial order redefined and retested', that approached the subject in an original way. He used a quasi-experiment drawing on techniques developed by cognitive psychologists for looking at serial order effects on memory and based on a series of telephone surveys between 1992 and 1994, conducted immediately after the Superbowl each year to investigate a number of clutter variables. In this study he sought to separate the effect of pure clutter (the number of other commercials in a break from serial order effects (where in a break the commercial ran). He also sought to investigate the impact of a number of advertisements for a given product in the same or nearby breaks. His findings were consistent with prior studies, showing a negative correlation between clutter and recall/recognition, with an estimated 38%--48% reduction in advertising effectiveness which may have been lost to cl utter competition. Serial order also seemed to be important, with earlier positions appearing to be more effective.
Zhao concluded that clutter might be thought of as having two components: proaction and retroaction (the effect of the number of other commercials preceding and succeeding respectively the test commercial in a given break). Their impacts were shown to be quite different, with proactive effects substantially stronger than retroactive effects, although the latter were still shown to be significant. Serial order effect could be defined as an exchange between retroaction and proaction. In addition, the magnitudes of the position effects demonstrated little variation across levels of frequency (the number of times a product was advertised in the same and neighbouring breaks). His analysis suggested that advertisers could compensate for lower frequency with more effective positioning. Getting rid of two to three preceding ads of other advertisers may have the same effect on brand memory as running one additional advertisement. Therefore, assuming equal price, advertisers should aim for fewer preceding and fewer su cceeding commercials. If the total number of ads in a break is set, the advertiser may want their commercial moved forward in a break. In other words, position may well be worth paying for.
In 1995, Krugman et al. had taken yet another tack using in-home observation studies in conjunction with the University of Georgia, in which they obtained data on viewer activities, commercial recall and 'eyes on screen'. Their study examined:
(1) The amount of visual attention to programming and commercials
(2) The relationship between activities while viewing and visual attention
(3) The relationship between visual attention to programming and commercials
(4) The relationship between attention to programmes and recall.
They found that programmes generated 66% eyes on screen time and commercials 33%, while visual attention was greater when viewers engaged in fewer activities. There was a strong correlation between 'attention' to programmes and 'attention' to commercials. The authors also concluded that programme involvement enhances commercial viewing. Finally, though, they report only some support for 'eyes on screen' during commercials correlating to recall and concluded that their methodology needed further development to address this test properly.
The 'First Impressions' Advertising Recall Study conducted by GMTV & Continental Research in 1998, based on a sample of 2,000 housewives, found amongst other things that commercial clutter was detrimental to the recall of commercials, and that attention to programming also led to improved recall of advertisements.
Kate Lynch and Horst Stipp took as their start point Erwin Ephron's 'The Myth of King Super Midas' paper at an ESOMAR/ARF symposium in Vienna in 1998, in their article 'Examination of qualitative viewing factors for optimal advertising strategies' (1999). In his paper, Ephron had put forward a modelling process that applied indicators of attentiveness to ratings information in order to build a 'probability of seeing a message' value for individual television programmes. That process considered eight factors:
(1) Viewers per set
(2) Ratings levels
(3) Appointment viewing
(4) Captive viewing
(6) Set location
(7) Programme type
Lynch & Stipp drew on this and a range of other studies and papers in their article to arrive at their conclusions, based on existing research, that viewing factors were likely to have a systematic, meaningful impact on commercial effectiveness. They are unlikely to have as much impact as the creative execution and the nature and quality of the audience, but they are important. They considered that programme appreciation ('appointment viewing', 'involvement' and 'loyalty') and the resulting higher levels of attention and tuning length to be the most important variables likely to impact on advertising effectiveness.
Erwin Ephron himself has sounded a cautionary note on some of the findings about the impact of quality of viewing measures in his follow-up 'Comments on the Stipp/Lynch paper' (1999). He asserts that most of the findings on programme liking as an influencer of commercial effectiveness are so far merely suggestive, and far from convincing. His view is that the real TV value issue is attention, not dial-switching, involvement or liking, and that this can be modelled from existing ratings information, based on such things as daypart, programme, who else is present, or which set is being used, as described above.
As recently as May this year, Jonathon Swallen of O&M presented the findings of two recent studies from 1998 and 1999 at an ESOMAR/ARF conference. These were that the more time people spend watching TV, the more likely they are to remember the commercials. Or in his words, 'If two viewers are exposed to the same ad, at the same time, in the same programme, the viewer watching more of the programme has a higher probability of ad recall'. He indicated that O&M were already starting to incorporate viewing duration factors into their TV planning and buying.
Quality of exposure studies in other media
Recent work in this area has by no means been limited to the area of TV viewing either. Back in 1990, TMP sought to address the age-old problem of standard press industry research, the fact that whilst Average Issue Readership provides a reasonable trading currency for agencies and media owners, it has been found lacking as a basis for planning advertising. This is primarily because of its failure to reflect how readers read a title, in terms of things like:
* Reading intensity
* Number of reading occasions
* Length of reading time (length)
* Reading location/situation.
In other words, those things which have been broadly described as reflecting 'quality' of reading.
Whilst the research did not specifically demonstrate the link between 'quality' of reading and advertising, it did identify the marked differences that exist between titles and audiences. The proportion of readers who were described as 'ideal readers' (people who read at least half of a publication and who described their reading as more than a quick glance through) ranged from 78% to 50%. It also showed that the impression a reader has of an advertisement is influenced by the publication in which it appears, showing the truth of Marshal McLuhan's famous dictum: 'the medium is the message.'
In the Netherlands, B. Walstra and P. Nelissen in 1992 took a different approach to understanding the impact of the print media environment on advertising performance. They tested how the performance of advertising in special interest magazines might be influenced by the interaction between the creative treatment and the nature of the editorial of the publication. They concluded that advertisements adapted to the media environment did have a greater impact than general, non-specific treatments.
Donthu et al. published the findings of their study 'Factors influencing recall of outdoor advertising' in 1993. The aim of the study was to capture the influence of seven factors on recall of outdoor advertisements:
(1) Position (LHS versus RHS of highway)
(2) Location (highway versus street)
(4) Number of words
(5) Respondents' involvement with the products being advertised
(6) Respondents' attention to outdoor advertising
(7) Respondents' attitude towards advertising.
In conclusion, using recall scores, the study found that the location, position, number of words, colour of outdoor advertisements, respondent involvement and attitude influenced the effectiveness of outdoor advertisements. Placing outdoor ads on the right-hand side (in right hand drive countries) on highways (rather than streets) may enhance their effectiveness. Effectiveness of outdoor ads will also be affected by the amount of respondent attention and their attitude toward advertising.
In 1994 Gruner & Jahr of the UK conducted a study into the 'presenter effect'. That is the impact that the magazine environment has on the communication of the advertising message. Their findings tallied with those of the Dutch study discussed above, showing differential performance between titles related to the nature of the magazines and their readership.
This was followed in 1996 by Louisa Ha's study, 'Advertising clutter in magazines -- dimensions & effects'. This was an experimental study to investigate the influence of clutter on magazine advertising effectiveness. Each level of response to advertising was covered. Three indices were developed to measure each dimension of clutter in magazines: quantity, competitiveness, and intrusiveness. The study hypothesised that all three dimensions of clutter negatively affect the attitudes towards advertising in a media vehicle and stimulate advertising avoidance, shown by a decrease in the general ad readership and a lowering of advertising message involvement.
The key finding was that individual differences in the perceived level of clutter vary more than expected and there is no common high-clutter level. Perceptions are therefore more important as predictors than actual levels of (say) intrusiveness. The worry of advertisers about the effect of clutter on advertising effectiveness is partially supported in this study. Both the quantity and intrusiveness dimensions of clutter lower the ad readership, while the intrusiveness dimension also reduces the memory of the focal ad among the subjects. It was suggested that the three dimensions of clutter identified in the study could serve as an additional evaluative measure of a media vehicle for an advertiser.
Around the same time, in the UK Heather Gould went back to a seminal piece of research into radio advertising effectiveness in her 'The ironing board revisited' study (1996). The original study had found that good levels of advertising recall were achieved with radio advertising, despite the low level of attention being paid to the medium. In this work, Heather replicated and extended the methodology of the original study to look also at the impact of different programme environments on radio advertising recall.
The findings in this update were similar to the original study on the first count. In addition, it was shown that solus spots contained within the news bulletins of local music stations (higher attention programme elements) delivered higher recall than other breaks.
Then, in 1997, a project was carried out in the UK designed to amplify the basic publication contact measures provided by a standard readership survey, by providing a comprehensive measure of repeat reading opportunities and a wide range of behavioural and attitudinal indicators. In addition, the opportunity was taken to assess the relationship between the readership of the main newspaper sections and the parent paper.
The Periodical Publishers Association initiated the project, known as the Quality of Reading Study (QRS), in conjunction with advertiser and agency organisations (1998). It was carried out as an independent study employing the main NRS readership survey methodology and classification data.
The main repeat exposure measure provided by the research, called PEX (Page Exposures) ranges from highs of 7.1 for women for bridal magazines, 6.1 for men for motoring performance magazines, with an average of 2.4 for all adults for all consumer paid-for titles. The behavioural and attitudinal measures collected by the study included five behavioural measures such as time spent reading and the number of pick-ups. In addition, agreement levels to eight attitudinal statements such as 'I expect to find lots to interest' and 'I read it when I'm relaxing' were collected. All these measures could be applied as media weights or simply used to enrich judgement about individual titles. A new QRS study has now been commissioned and the results will be available in the Autumn of this year.
It is clear that the quest for a better understanding of the contribution which the quality of media exposure makes to the effectiveness of advertising has been a long and wide one.
Early on there was strong evidence that attention and clutter effects were at the heart of this issue. Much of the work carried out in more recent years has been about trying to find better ways of measuring factors such as attention, position effects and involvement to answer the two key questions:
* How do we determine a better measure of who has actually looked at a specific advertisement placed within a medium?
* How do we determine which people are in an appropriate frame of mind to respond to the advertisement?
These later studies have sought to find ways of applying our knowledge and understanding to improving the planning of advertising media strategies. What seems to be beyond doubt is the importance of these quality factors that are beyond the OTS.
However, as the debates go on it also seems that our knowledge and understanding is not yet complete. We do not yet have definitive ways of taking and applying the insight we do have to the planning process on an ongoing basis. Progress has been made in this direction, and, like Carat, many advertisers, agencies and media owners are now trying to take account of quality of exposure factors to improve the effectiveness of advertising campaigns. Indeed, it would be foolhardy for anyone to ignore all this learning in arriving at media decisions today. However, it will be interesting to see how much further our knowledge, understanding and application advances over the next 50 years.
James Galpin is International Projects Director at the Carat Insight in London, the marketing communication consultancy arm of the Carat Group. He has twelve years' experience of media, research and marketing consultancy, working for a range of blue chip clients.
Phil Gullen is currently the Managing Director of Carat Insight. Prior to this he headed Carat Research and was a board director at J W Thompson, where he was responsible for both R&D and media planning. He is a Fellow of the IPA and a former Chairman of the Media Research Group.
Advertising Research Foundation (ARF) (1956) A Study of Printed Advertising Rating Methods. New York: ARF.
Billett, J. (1993) Share of Break Study. London: The Billett Consultancy.
Billett, J. (1998) Commercial Break Ecology. London: The Billett Consultancy.
Carat Research (1993) Foretel Advertising Recall Study. London: Carat Research.
Carat Research (1994-9) Foretel Attention Study Programme. London: Carat Research.
Clancy, Shulman & Associates (1984) Program Impact & Program Appeal: Qualitative Ratings & Commercial Effectiveness.
Consumer Dynamics Corp. (1980) A Study on the Effects of 'Clutter' in Magazine Advertising.
Continental Research (1998) 'First Impressions' Advertising Recall Study. London: GMTV.
Donthu, N., Cherian, J. & Bhargavia, M. (1993) Factors influencing recall of outdoor advertising. Journal of Advertising Research, 33, 3, pp. 64-72.
Ephron, E. (1998) The myth of King Super Midas. ESOMAR/ARF Worldwide Electronic & Broadcast Audience Symposium. Vienna: ESOMAR.
Ephron, E. (1999) Comments on the Lynch/Stipp paper. Journal of Advertising Research, 39, 3, pp. 17-19.
Gould, H. (1996) The ironing board revisited. Admap, April.
Gruner & Jahr (1994) Unravelling the Message. London: Gruner & Jahr.
Gullen, P. (1990) Women beyond 'AIR'. Admap, November.
Ha, L. (1996) Advertising clutter in consumer magazines -- dimensions & effects. Journal of Advertising Research, 36, 4, pp. 76-84.
Krugman, D., Cameron, G. & White, G.M. (1995) Visual attention & commercials: the use of in-home observations. Journal of Advertising, 24, 1, pp. 1-12.
Lieberman Research (1986) The Impact of Editorial Environment on Brand Acceptance -- A 13 Magazine Analysis. Los Angeles: Lieberman Research.
Lloyd, D.W, & Clancy, K.J. (1995) CPMs versus CPMIs: implications for media planning. Journal of Advertising Research, 35, 4, pp. 34-44.
Lynch, K. & Stipp, H. (1999) Examination of qualitative viewing factors for optimal advertising strategies. Journal of Advertising Research, 39, 3, pp. 7-16.
Newsweek Inc. (Copyright (c) 1977, 1984 & 1991) Newsweek Media Research Index.
Ogilvy, Benson & Mather (1965) Experimental Study of the Relative Effectiveness of Three Television Dayparts.
Periodical Publishers Association (1998) Quality of Reading Survey. London: Periodical Publishers Association.
Ray, M. & Webb, p. (1979) Effects of TV clutter. Journal of Advertising Research, 19, 3.
Ryan, T.E. (1968) A Study of Reading Environment and Advertising Memorability.
Simmons, W.R. (1968) Pilot Study on Immediate Recall of Television Advertising.
Swallen, J. (2000) Viewing stamina boosts recall. ESOMAR/ARF Conference. Bal Harbour: ESOMAR.
Walstra, B. & Nelissen, P. (1992) Adapting advertising to the media environment. ESOMAR Media Research Meets the Future Seminar. Lisbon: ESOMAR.
World Advertising Research Centre (WARC) (2000) www.WARC.com.
Young, E. (1983) Eye Tracking: One Approach to Measurement of Editorial Environment. ARF Magazine Research Workshop. New York: ARF.
Yuspeh, S. & J. Walter Thompson (1977) On-air: Are we Testing the Message or the Medium?
Zhao, X. (1997) Clutter and serial order redefined and retested. Journal of Advertising Research, 37, 5.
Carat Foretel Ad Recall Study. Centre break of Home and Away Time Commercial Duration Correct recall (%) 18:12 Cadbury's Boost 20" 18 18:12 Kraft Golden Crown 30" 14 18:12 Royal Doulton Figurines 30" 9 18:13 Walker's Crisps 30" 12 18:13 Kellogg's Frosties 20" 11 18:14 Allied Carpets 10" 9 18:14 Quality Street 30" 16 Break average: 13 Source: Carat Foretel Advertising Recall Study/Broadcast Research % of adult viewers paying full attention 1992 1994 1996 1998 (%) (%) (%) (%) ITV 46 41 38 42 Channel 4 43 43 45 46 TV-AM/GMTV 12 21 14 22 UK Gold 33 33 39 31 The Programme effect ITV one-hour dramas at 9 p.m. on Tuesdays Boon Soldier Soldier (Nov 1992) (Nov 1993) Specially chose (%) 48 69 Enjoyed (%) 71 77 Paid attention (%) 68 83 BARB audience (millions) 11 13 Soldier Soldier (Nov 1994) Specially chose (%) 80 Enjoyed (%) 82 Paid attention (%) 92 BARB audience (millions) 15 Source: Carat Foretel Attention Study, BMRB, BARB Viewer Involvement Profile: two game shows--all adults Fifteen to One Lose a Million (%) (%) All viewers 100 100 Magic moment 53 5 Regular date 30 19 Change meeting 9 60 Wallpaper TV 8 16 % of new car buyers for whom the programme was a 'magic moment' Test Match (cricket) 54 Heartbeat (drama) 43 Fifteen to One (quiz) 40 Coronation Street (soap) 30 This Morning (daytime TV) 8 MTV 2 Talking Telephone Numbers 0 VIP versus Attention Attention level Full Average Little None Total Magic moments 86 13 1 0 100 Regular dates 49 45 6 0 100 Chance meetings 18 46 32 4 100 Wallpaper TV 1 3 65 31 100 Total 40 30 23 7 100 Commercial break behavior (All present in the room with the set switched on to a channel) Actual % Re-based % Saw all the break 32 59 Stayed and talked 7 41 Stayed and did something else 15 54 100 Effectiveness by position in break Proportion of OTS correctly recalled against the average Full database (12,125 observations) First -9% Middle 2% Last -1% Source: Carat Foretel Ad Recall Study/Broadcast Research Effectiveness by break size Proportion of OTS correctly recalled against the average 1/2 55% 3/4/5 19% 6/7/8 -3% 9+ -3% Source: Carat Foretel Ad Recall Study/Broadcast Research Effectiveness by programme attention Proportion of OTS correctly recalled Full 18.6% Average 17.5% Little 12.3% No 7.6% Source: Carat Foretel Ad RecallStudy/Broadcast Research
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||opportunity to see|
|Author:||Galpin, James; Gullen, Phil|
|Publication:||International Journal of Market Research|
|Article Type:||Statistical Data Included|
|Date:||Dec 22, 2000|
|Previous Article:||The future of multimedia research.|
|Next Article:||International market research.|