Correlations between reach and circulation can best be determined for similar print titlesBy Wolfgang Koschnick No sooner had the advertising industry obtained reliable data on the paid circulation of print media with a functioning circulation certification than the first doubts arose: What use are circulation figures? What matters is the reach. But these develop very differently depending on the print title. Mathematical models are also of little use when it comes to determining clear correlations between the two values.
Until the 1950s, circulation figures were the only data available on printed media. However, they provided at best indirect information about usage behavior, media contacts and reader characteristics. Circulation figures lost their importance for media planning in Switzerland from the mid-sixties at the latest, after Wemf AG für Werbemedienforschung provided the first readership figures based on sample surveys (Mach Basic).
And now, like every year, the media landscape is buzzing - the same thing happens again and again: a magazine has seen its paid circulation soar, its advertising business is developing splendidly, and then Mach Basic confirms that its reach is declining or is bobbing along. No publisher can or wants to believe this. The figures are lazy and Wemf is manipulating them, they complain. But the truth is that there are quite reasonable explanations for this phenomenon.
The strong differences between circulation figures and reach figures first caught the eye of British media researchers. The pattern is often similar: first, both figures rise sharply. Then the reach figures stagnate or even fall, while the circulation figures continue to rise. To clarify this, the two researchers T. Corlett and D. W. Osborn were commissioned by the British Institute of Practitioners in Advertising (IPA) in 1966 to carry out the world's first study on the sharp differences in reader-per-copy values (LpE) for comparable publications.
They developed the hypothesis that the differences between the titles were due to the different composition of the editorial content and the four-color section of a Sunday newspaper, the Sunday Telegraph. The four-color printing attracted more readers per copy, but not more buyers - a finding that does not explain the general phenomenon at all.
In another study in 1970, John Bermingham, R. Gage and P. Mills were able to show that when newspaper circulation decreases, there is only a relatively small decrease in LpE values, while a significant increase in circulation is expected to lead to a considerable decrease in LpE values.
In an article entitled "Magazine Readers per Copy", Timothy Joyce then emphasized in 1974 that the LpE was only an average value. It is possible, for example, that 20 percent of copies account for 50 percent of readers. According to Joyce, two factors in particular influence the LpE value: the demand for reading and the availability of copies. Availability on public display increases the value, whereas promotional activities to attract subscribers reduce it. Consequently, the LpE will never be a stable figure.
At the hairdresser, the number of readers skyrockets
Further studies show that the relationship between circulation and reach depends on various factors: A "Report on the relationship between circulation and LpA values" stated: "And it is safe to assume that our previous idea of the relationship between reach and circulation will have to be corrected; there is certainly no uniform relationship, the relationship appears to be 'title-related' based on the structure of the readership."
The study was written by Erich Ockelmann for the Technical Commission of the German Arbeitsgemeinschaft Media-Analyse (AG.MA) in 1979. Ockelmann came to the conclusion that the underestimation of circulation occurs when the readership consists of older people, women and people with below-average magazine contacts.
In 1985, Michael Brown developed a mathematical model that reflects the relationships between circulation and reach for different countries, measurement methods, publications and points in time. His analysis showed that these relationships are not linear. He explained this as follows:
- If the circulation is increased, the number of opportunities for second readings increases proportionally.
- But the proportion of these
The number of new opportunities that actually lead to readership is decreasing for two reasons:
Firstly, the probability increases that people who come across a copy by chance are already readers. This means that the number of people who have not yet read the magazine decreases.
Secondly, an increasing circulation could mean that the remaining non-readers have less and less interest in this publication.
Consequently, according to Brown, it is inevitable that each added copy attracts fewer readers than the previous one.
Finally, two studies carried out in 1987 by the British Joint Industry Committee for National Readership Surveys showed that even very high readership figures can be correct.
The first of these two studies was based on the characteristics of three women's hair fashion magazines, with readers per copy varying from 2.5 (for a new title) to 13. This study examined on a smaller scale the number of magazines available in hairdressing salons and the extent to which they were read. It was found that the LpE value for the new title was too small, while the other readership figures were fairly accurate.
It was estimated that a magazine in such a location could easily attract 200 to 300 readers over the course of its "lifetime". As a result, this high number of readers of some copies drives the total LpE figure to a very high level. However, the study also showed that more than half of all reading of these magazines took place at the hairdresser.
A subsequent study on overall reading behavior outside the home found that the
The fact that the hairdressing industry generates 500 million contacts with magazines that are available there.
Program guides and Sunday supplements, on the other hand, are hardly read outside the home according to this study. But for monthly titles, out-of-home reading still accounts for a sixth of the total number of readers in the National Readership Survey (NRS). Reading outside the home is even more important for bi-monthly titles and quarterly magazines; the proportion of out-of-home reading here is no less than 62 percent.
These analyses point to the connection between circulation size and publication interval on the one hand and the importance of out-of-home reading on the other: a bi-monthly or quarterly title can collect readers over a very long period of time. Many of these titles are physically robust, so they have a long life in public places. The editorial content of these magazines is very suitable for reading in such places.
To summarize: Primary reading, i.e. reading by the buyer and their immediate family, is a function of circulation and should largely follow circulation trends. However, this does not apply to secondary reading and certainly not to secondary reading of out-of-circulation magazines.
Price, distribution and circulation
influence the LpE value
Criticism of the LpE value is primarily directed against the
Survey method in user structure analyses. This is based on "hearsay" information: The surveyed reader is asked to provide information on how many people in total read the copy of the magazine or newspaper they use. This value is then used to extrapolate the reach (readers per issue).
However, a sufficiently accurate reach value can only be calculated if it is possible to identify all potential readers of a newspaper or magazine.
to be included in the sample and asked about their use of the specified edition.
The German readership researcher Eva-Maria Hess wrote: "The credibility of reach analyses is often measured by readers-per-copy figures, because these figures represent the bridge to the objectively ascertainable circulation figures. An excessively high or low value for 'readers per copy' always triggers considerations as to the extent to which systematic errors may have occurred in the title in question. There are genre-related
which influence the LpE values and which are
should be taken into account when considering plausibility. Such characteristics include, for example, the forms of distribution, the copy price and the absolute circulation."
Several explanations can be given for the decline in many LpE values:
- The reading circles have lost much of their importance, which means that repeated borrowing
of magazines against payment.
- The average household size has decreased.
- The magazine market has expanded considerably, with a growing number of special interest titles in particular.
In other words, the number of readers in the household is declining and therefore the reach of an average issue, i.e. the number of readers per issue (LpA), is falling, even if circulation increases or remains constant.
In a ten-year comparison from 1979 to 1989 of six national daily newspapers and 25 magazines in German-speaking Switzerland and 14 daily newspapers, two weekly newspapers and ten magazines in French-speaking Switzerland, the LpE values for magazines fell sharply from 2.49 (1979) to 1.79 (1989). The decrease amounted to 0.07 LpE units per year.
In contrast, the national daily newspapers showed a mixed picture: While readership figures held steady on average, circulation figures showed an upward trend. This also resulted in a decrease in LpE values by an average of 0.03 units per year.
The free titles occupied a special position. While they still had 1.2 readers per copy in 1979, only 1.07 LpE remained ten years later. The situation was the same in both German-speaking and French-speaking Switzerland - with the exception of the daily newspapers in French-speaking Switzerland, which even managed to gain readers.
In 1996, Hans-Erdmann Scheler summarized the previous studies and their findings as follows: "If a title is new on the market and is really attractive for a target group, it first collects readers, i.e. one copy has many readers. If the title remains attractive, buyers (purchasers and, less frequently, subscribers) are recruited from among the readers, with the result that circulation increases. If a title becomes less attractive, the readers turn away from it first - with the result that the reach decreases. Only when the title becomes so unattractive to purchasers that they no longer want to read the magazine does circulation decline."
Media researchers also have to live with a black box
Finally, the Viennese media researcher Karl Steurer developed the hypothesis from earlier studies and more recent Austrian data in 1999 that only titles with an identical publication frequency, identical circulation structure and identical magazine type could - but did not have to - be expected to have a similar readership size. Steurer offered several explanations for this:
- The longer the publication period, the greater the chance of gathering potential readers - the widest readership (WLK) - and the greater the chance of generating readers from it (LpA). If two magazines are certified as having the same circulation and one of them is a weekly magazine and the other a monthly magazine, the readership (WLK and LpA) will rarely be the same.
- The circulation structure has an influence on the readership that can be reached. Newspapers with the same circulation, but with different circulation structures, therefore produce different readerships.
different readership figures. The readership of a newspaper that consists almost exclusively of subscribers will again arrive at different figures than that of a newspaper that is mainly sold individually or free of charge.
- The marketing target group must also be taken into account. As a rule, every publisher will have an idea of who its buyers (readers) should be, more or less confirmed by market studies. If the marketing target groups of two properties are not the same, it cannot be expected that the readerships will be the same under the same circumstances.
Based on this data, Steurer came to the following conclusions:
- Only if the influencing factors for two objects are largely identical can it be expected that the same print runs will generate a similar number of readers. However, even these influencing variables are not sufficient to satisfactorily explain readership figures from a given circulation.
- Circulation alone explains "only" 76 percent of the variance - deviation from the mean - of the LpA figure. Although this is a pretty good value in itself, it is still not good enough. But at least
the circulation is the main influencing variable.
- If, in addition to circulation, the circulation structure is added as an explanatory factor, the explained variance increases to 82 percent. However, even then, 18 percent unexplained variance remains. This means that neither the circulation nor its structure are sufficient to explain or even predict readership figures.
There are more and more overzealous studies and ever greater perplexity in the communications industry, but somehow you still can't shake off the feeling that the connection between circulation and reach has not yet been properly discovered. No matter how you look at it, there is no direct correlation between the reach of a title and the number of copies it sells. At the same time, it seems obvious that the two elements are connected.
Smaller households push the LpE value down
What is certain is that the mathematical-statistical approaches do not represent a solution: Even if you take Michael Brown's logarithm seriously, it says nothing about the content of the relationship between circulation and reach. It merely defines the boundaries within which the relationship takes place.
In any case, examining the relationship only makes sense for similar magazines. There are magazines that have very many readers per copy and those that have very few. As a rule, this is related to the editorial content and lifespan on the one hand and the general living conditions of the recipients on the other.
Magazines whose editorial content is read for months on end and can be read over such long periods of time due to their overall design naturally have more readers per copy than those whose content is outdated after a few days. And if families are getting smaller and smaller and more and more people are living as singles, the number of LpEs will naturally also fall. Mathematizing these correlations excessively is unproductive because the figures are less meaningful than the verbal explanation of the correlations.
In the case of similar publications, such as various automobile magazines or magazines for home and garden, it can be seen with a
The higher LpE values reflect the differing appeal of the various titles as an expression of different editorial concepts. In this respect, higher reach with the same circulation in a comparison of titles can - with considerable reservations - be seen as an indicator of higher editorial quality.
Basic arithmetic operations for reader researchersWhile the reach (LpA - readers per issue) expresses how many people use an average issue of a title, the LpE value indicates how many people on average use a copy of the issues of a title. This figure is not calculated, but is the result of an arithmetic operation.
The credibility of representatively surveyed reach is often measured by LpE values because these figures represent the bridge to the circulation figures determined by secondary statistics. However, there are genre-specific features that influence the LpE values and should be taken into account when considering plausibility. These include the distribution types, the magazine price and the absolute circulation figures.
The number of readers per copy of a newspaper or magazine varies between 1 and 7 readers, depending on the type of magazine or newspaper, and reaches a higher value for only a few titles. The LpE value is therefore the sum of the first and subsequent readers or the primary and secondary readers. For some time, the number of readers per copy calculated from the number of copies distributed or sold and the total number of readers was considered a useful measure of the reach of an advertising medium. (wjk)