Sections research: a new way forward?

The author presents the findings of Sections 95, the weekend section readership research conducted for Express Newspapers, Leo Burnett and the COI.

Sections research: a new way forward?

Jennie Beck presents the findings of Sections 95, the weekend section readership research conducted for Express Newspapers, Leo Burnett and the COI by Beck Consultancy

Jennie Beck
IT IS OVER 30 YEARS since the first spawning of a new section in a national newspaper, and over ten years since we saw the majority of weekend quality papers coming in multi-sectioned format. That was when the demand for section research really Of course, during this time, independent research initiatives have shed light on the issue - from the Quality Sunday survey in 1988, right through to this year's CIA Medialab work - and there have been private surveys too. But there was no comprehensive sections research, funded by all sides of the industry, until Express Newspapers, Leo Burnett and the COI announced Sections 95 last July. Before the project was announced, it was discussed with leading industry figures, and comments sought on the survey design and questionnaire. As a result of this careful consultation and planning, Sections 95 has been granted a starred appraisal from the IPA. What did we do and how did we do it? First, as the vast majority of section expenditure decisions are focused on weekend issues, we looked only at Saturday and Sunday quality and mid-market papers. Thus, we covered 12 titles, carrying, between them, a total of 59 sections - a section being any separately printed and titled part of a paper. We decided not to cover those so-called sections that appear run-of-paper, partly because of the unpredictability of their appearance or positioning, but mostly because readers tend not to regard them as separate sections and there is, therefore, considerable potential for respondent confusion. Our 59 sections included 12 main news sections, 11 colour tabloids, four colour magazines, two magazines that are really colour tabloids, two review sections that are not colour tabloids, four sections with review in the title, that are not all review sections, and five sections called 'Weekend'! When it came to designing the survey, the primary objective was a face-to-face interview with a specific issue reader, using the specific issue as a prompt. Thus, in order to conduct the interview in-home - the ideal scenario - we needed to arrange fieldwork to overcome the practical problems of lugging 59 sections from door-to-door, and to give the interviewer as straightforward a task as possible. We split the 12 titles into four groups, on the basis of similar readership profiles, and, in effect, conducted four identical surveys simultaneously. The groups were: Sections 95 was conducted among a quota sample of ABC1C2 specific issue readers of the 12 titles, in England and Wales. After some debate, we decided to exclude Scotland from the survey because of the differences in the newspaper market and in editions of national titles north of the border. Each of our four survey groups was treated separately when it came to selection of sampling points and setting of quotas for age, sex and social class. NRS July-December 1994 data were used to distribute the sample by ISBA region and to set the quotas. Each title's results were subsequently rim-weighted, by age, sex, class and working status. The survey interviewed a total of 1,982 specific issue readers with individual title samples all over 130. Interviewing was conducted face-to-face, in-home, with the interviewer carrying copies of all the papers in his or her group. The issues we used were those for Saturday 15 July and Sunday 16 July. Although we could not have predicted it, in many ways this was an ideal weekend to choose. There were no special one-off sections published and, although the chose that weekend to launch the 1015 section, we had sufficient warning to add it to the questionnaire. There were no distribution or availability problems that weekend and no unusually high profile news stories. I should say, sadly, there were no unusually high profile news stories because the main news that weekend was, once again, Bosnia, with all the Saturday papers and all the Sunday qualities leading on the refugee crisis following the abandonment of the safe havens. These actual issues, with all their component parts, were used as prompts during the interview, which took between 20 and 40 minutes, depending on the size of the paper. Interviews were conducted in 80 sampling points in England and Wales, and all took place between the Monday and Friday following the publication of the specific issues being surveyed. Four recruitment questionnaires were used, one for each group, to screen for papers normally read and specific issue readership, as well as to collect demographic data and check quotas. Then an individual title questionnaire was used for the main interview. Although 12 separate questionnaires were produced, they carried common questions and common wording. Only the titles and section descriptions were changed. Exhibit 2 shows the range of quality of reading questions that were asked about normal reading behaviour, about reading of the specific issue, about the paper as a whole, and about individual sections. Collecting normal reading frequency data for the paper and for an individual section allows us to compare the two and, thus, to calculate a normal (or claimed) section patronage figure - equivalent to the percentage of a parent title's readers who will read or look at something in the section. We also, of course, asked about patronage of the section in the actual issue - showing the section cover as a prompt. This allows us to compare what people say they do and what they did that weekend. Other quality of reading measures covered in the questionnaire included amount read, time spent reading, two different reading style questions, disappointment if not available, source of copy and overall impression of the sections. Although most questions were asked about all sections, some of the quality of reading measures were dropped for the minority sections - for example, appointments and cartoon sections - that do not carry display advertising. In addition to all this, questions on the reading of editorial subject areas and order of reading were included. All this has produced a huge volume of individual title data, as well as opportunities for combining data across demographic subgroups, or across similar sections or titles. From this cornucopia, I have picked out a couple of examples that illustrate the value of section data and, I hope, shed some light on the complexity of the weekend newspaper market. Exhibit 3 shows actual section patronage figures for all review sections - that is, colour tabloid sections which are, confusingly, known as reviews - and for all colour magazines. These figures range from 77 per cent, for one of the two colour tabloids, up to 90 per cent plus. What is interesting about this analysis is that the introduction of review sections has not damaged patronage of the colour magazines that accompany those sections. The magazine is considerably better patronised than its two review sections, and the and magazine figures are close to those of their respective review sections. The lowest magazine patronage score is for the , even though, apart from the TV guide, it is the only colour element in the package. The average section patronage for review sections, at 88 per cent, is only marginally higher than the 87 per cent for colour magazines. The main difference between the two formats comes in average time spent reading, shown in Exhibit 4. This shows average minutes spent reading by viewers of each review section or magazine. Here, the colour magazines are further down the list, with the magazine dropping from seventh place to twelfth, and the magazine dropping below the review sections. This helps bring the average reading time for all magazines down to 19 minutes, compared with 22 minutes for all review sections. Women tend to read more and spend longer on the review sections and magazines, although they could not be described as definitely female sections - Sections 95 was conducted before started positioning itself as a women's magazine. There are, however, weekend sections aimed particularly at men, so do they read more in these 'male' sections? Here are a couple of examples. Exhibit 5 shows that male sections would appear to be quite well defined in the . Men read more on average (the first column) in Business News than they do in Weekend, despite the fact that their patronage of the latter is slightly higher. Similarly, men read more of Motoring than they do of the Magazine, although the latter's patronage is rather higher. A similar comparison for the, before the recent relaunch, says a lot about the weekend preoccupations of its readers. Sport is a very successful male section, with men reading over 50 per cent on average, but Business comes bottom of the list after Review and Life, with male readers reading only a third of the section. I believe Sections 95 holds three key lessons for an industry research solution to the sections problem. First, given the high readership of review sections, it makes no sense to have some of them on the NRS and some not. It makes no sense to have magazines on the survey but not reviews. But adding review sections to the questionnaire - the current plan - just adds to the potential for title confusion, particularly for infrequent readers, and gives no indication of readership of other key advertising sections. Indeed, I would go further and suggest that everything the parent title should come off the survey. Sections research should, in my view, be conducted off-survey. Secondly, a useful by-product of this research is the opportunity to examine the discriminating power of a raft of quality of reading measures. The most valuable measures - those that are consistent discriminators - proved to be 'disappointment if not available' and 'amount read'. Those who would be very disappointed if a paper was not available tend to read more sections, read more in them and spend longer on them than other readers. The same is true for those who normally read all or almost all of a paper. Less reliable indicators were source of copy and reading frequency. Most importantly, this survey confirmed the widely-held belief that some form of specific issue or actual section prompt is needed to ensure accurate respondent recall. Mastheads or titles with short descriptions will not do. This is not true for all sections, of course. Exhibit 7 shows the comparison between normal section patronage - calculated from normal reading frequency for paper and section - and actual section patronage for grouped sections. For the better-read sections, there is very little difference between claimed and actual section reading. However, it is clear that some readers are under-reporting their readership of the minority sections, until they actually see them. So, are specific issue sweeps the way forward? The idea of surveying a handful of titles at a time has a number of advantages: it allows use of specific issues for in-home interviews and it is flexible enough to accommodate change. It could be the answer to the difficulties at least and, if repeated on a regular basis, specific issue work will also provide valuable insight into the effect of content on readership patterns.

WARC subscribers can sign in to keep reading

Not a subscriber?

WARC helps you to plan, create and deliver more effective marketing

  • Prove your case and back-up your idea

  • Get expert guidance on strategic challenges

  • Tackle current and emerging marketing themes

WARC consistently delivers valuable insights that help to make me look like an expert in front of my colleagues and clients.

Dasha Boryso
Strategy Partner, Fetch

You’re in good company

We work with 80% of Forbes' most valuable brands* and 80% of the world's top top-of-the-class agencies.

* Top 10 brands