Repeat reading
Magazines are not just read once; frequently they are read on several occasions, and many pages (including advertisements) are looked at repeatedly. This reflects one of the benefits of print media over television and radio - the readers’ ability to control the timing of their exposure. The desire to re-visit a magazine is an expression of the relationship between reader and magazine.
In picking up a magazine several times, each reading may well cover only part of the total pages (except perhaps the first reading), but by the time the reader has finished with the issue virtually all of the pages will have been seen, some of them a number of times.
The pattern of repeat reading is most easily observed through a diary panel. IPC’s “Media Values” research included a diary panel of 250 adults [36] recruited from the main Media Values survey. Panel members kept a diary of their magazine reading for two weeks. There were numerous instances of separate reading occasions during the course of a single day. Examples are:
Woman aged 34, on a Monday: reading of TV Times
11 am 15 minutes, in living room, no other activity.
6 pm 10 minutes, in kitchen, no other activity.
9 pm 15 minutes, in living room, while watching TV.
She had read TV Times for a total of 40 minutes that day, in three separate sessions. In addition she read Bella for 20 minutes this Monday.
Man aged 32, on a Sunday: reading of Arena
10 am 15 minutes, in lounge, no other activity.
Noon 15 minutes, in lounge, while listening to radio.
Man aged 28, on a Thursday: reading of Shoot
9 am 30 minutes, in lounge, while listening to radio.
Noon 10 minutes, in lounge, while eating.
4 pm 10 minutes, in lounge, no other activity.
8 pm 15 minutes, in lounge, while watching TV.
This totalled 65 minutes on one day.
This enables us to see examples of people reading a magazine on two, three or four separate occasions within a single day, adding up to anything from 30 minutes to more than an hour. In addition these issues may have been read on other days.
In order to look at the incidence of repeat reading of the same issues on different days I turned to an earlier diary panel and examined individual panel members’ records through time. This was an experimental panel commissioned in 1984 by JICNARS [37], and run as three separate sub-panels by three different research companies (Communications Research Ltd, Research Bureau Ltd, and AGB Cable & Viewdata). The panel ran for about four weeks during October and November 1984.
There were countless instances of reading an issue of a magazine on more than one day. A few examples will be illuminating. The accompanying table shows extracts from the diaries of four panel members.
Mrs F is typical in having a number of instances of reading a magazine over two or more days. The 3rd November issue of Woman’s Weekly was someone else’s copy; Mrs F began reading it for the first time on 16th November when it was nearly three weeks old. She read it in her own home. Next day she read the same issue again, at home. She also read the 10th November issue of Woman’s Weekly on those two days. It looks as though a friend or relative gave both issues to her on the same day, and she was reading them in parallel. On 19th November Mrs F began reading the new issue of Woman, cover-dated 25th November. It was her own copy, she read it at home, and she read it again next day.
Mrs W read the new copy of Family Circle on two consecutive days, 26th and 27th October, then after a gap of two weeks she read it again, and five days later read it on the fourth different day. It was her own copy and she read it at home each time.
Mr G read the household copy of Radio Times every day for seven consecutive days. He did the same with TV Times. Every week there was this same pattern of reading both magazines each day - obviously for planning his daily viewing.
Mr E was an enthusiast of Reader’s Digest. He read the October issue on 21st October; it was not the first time he’d read that issue, so he had evidently started reading it before he began keeping his diary that day. He also read that issue on four other days over the next two weeks. It was a household copy of the magazine, and each time it was read at home. He acquired the November issue and read it for the first time on 8th November, at home. He then read it on seven of the next nine days.
And so it goes on. The diaries are filled with instance after instance of magazine issues being read on more than one day, sometimes two or three weeks apart, besides other issues being read on one day only.
Key:
- First time (i.e. is this the first time of reading this issue?): Y=Yes, N=No.
- Whose copy: O=Own, F=Family/household copy, S= Someone else’s.
- Where read: H=Own home, S=Somewhere else.
The diary did not attempt to measure the number of pick-ups within a single day. However the 2000 "Quality of Reading Survey" [33] measured the average number of pick-ups, which reflect both the pick-ups within a single day and the different days on which an issue was read. The question asked “How many times do you usually pick up an issue of ……… by the time you've finished with it?” The resulting averages are shown below for a small selection of the magazine categories:
|
TV listings weeklies |
Average number of times picked-up 5.4 10.0 |
The average copy of a paid-for magazine is picked up 5.4 times by adults. TV listings weeklies understandably have the highest number of pick-ups because of their reference use, but among the 31 categories of paid-for magazines even the lowest-scoring category achieves 3.6 pick-ups of a typical copy.
It is clear that repeat reading of magazines is a major benefit for advertisers, yet it is one that is not reflected in the National Readership Survey figures.
Page EXposures (PEX)
The previous section has shown that magazine issues are often read on more than one day, that they may be picked up and read more than once within a day, and that more than one issue may be read on a single day. It is also true that not all of the issue is necessarily read on a day when it is picked up.
These factors were combined into a single score - PEX (Page EXposures) - in the 2000 "Quality of Reading Survey" (QRS) [33]. PEX measures of the number of times the average page in a magazine is read or looked at by the average reader. In effect, this means the average number of times a reader will see a typical advertisement.
PEX was introduced because the National Readership Survey treats all magazines as offering equal advertisement exposure, even though this does not reflect reality, and because magazines offer more impacts than the average issue readership figures allow, as the previous section has indicated. With magazines these extra impacts are free, unlike the broadcast media where every transmission costs extra.
The concept behind PEX is simple. First, establish the average number of different days on which a magazine is read (within its publishing interval), and also the average proportion of pages opened on a typical day when it is read. Then multiply the two together to obtain the total proportion of pages opened - which converts to the number of times an average page is looked at.
For example, suppose that in a typical week a given weekly magazine is read on three different days, as an average across all readers; and that on a day when the magazine is read an average of 60% of the pages are opened. It is easy to see that during the whole week 3 x 60% of the pages are opened, i.e. 180%. This means an average of 1.8 times per page.
In practice the questions which measure PEX ask about the number of reading days in the last week/month/etc, rather than the average week/month/etc; and the proportion of pages opened on the last day rather than the average day. This is because it is easier for respondents to answer in terms of a particular recent occasion than to estimate an average over a longer period. But it can be demonstrated from QRS data that the average derived from a large number of people reporting on the last occasion produces the same result as the average derived from the same people reporting on their individual average behaviour. PEX also uses an additional question to take account of people sometimes reading more than one issue on the same day.
Thus the three questions that make up the PEX score ask:
- The number of different days on which any issues of a named magazine were read or looked at, within the publishing interval (e.g. 7 days for a weekly)
- The number of separate issues read or looked at on the last day on which any issues were read
- The proportion of pages opened on the last day on which an issue was read (the proportion of the last issue read if two or more issues were read that day)
The PEX scores are calculated by multiplying the three answers together.
The average PEX score across all the paid-for magazines covered by QRS was 2.54. In other words, the average magazine page is looked at 2.54 times, a great benefit for advertisers. For national newspaper sections and supplements the average PEX score was 1.28. There were considerable variations by publication categories, as the ranking of the scores on the next page indicates.
No. of times the average page in a magazine is read or looked at
Source: QRS 2000
The adult page exposure scores range from 4.86 for performance car magazines to 1.56 for motoring customer magazines, and 1.28 for newspaper supplements/sections. The reason for this range is the different ways in which different categories of magazine are used by their readers. There are further variations between titles within each sector. And different kinds of reader are liable to show different scores: the general principle is that core readers have rather higher page exposures than non-core readers.
All categories of magazine deliver a higher exposure to advertisements than NRS average issue readership suggests, for the NRS implies a PEX score of 1.00 for all titles.
To make it more practical for media planners, buyers and sellers to take account of PEX and other QRS data when planning print media schedules, the QRS sponsors have arranged for the QRS database to be fused onto the new NRS database every six months, and held by the computer bureaux alongside the un-fused official NRS database.
In case it might be thought that the repeat exposures have no value for advertisers, there are three classic studies carried out by Alfred Politz in the USA [38]. For each study, 12 advertisements were tested among three matched samples of subscribers to a magazine.
One sample was not exposed to the test advertisements at all, another was exposed to them once, and the third was exposed twice. Four different measures (brand familiarity, claim acceptance, rating of brand quality, and interest in buying) all gave the same answer: compared with the control sample who saw none of the test advertisements, the effect on the people who saw the test advertisements twice was roughly twice as great as the effect on those who saw them once. That is, the additional effect of a second exposure in these magazines was roughly the same as the effect of the first exposure. Repeat exposure matters.