Readership and circulation: reasons why they might diverge

A question that is likely to be asked of The Readership Works is how can its figures for a particular title move in a different direction, or to a greater or lesser extent, than changes in the ABC figure for the same title? The perplexity this causes will be understandable, but there are, in fact, legitimate reasons why readership and circulation might diverge.

The relationship between readership and circulation is a complex one. This can be seen in the large variation in the number of Readers-per-Copy (RPC) between different titles on the TRW. RPC is the average issue readership of a title divided by its circulation. RPC is not static: there can be long-term or short-term changes to the RPC of a title, and a variety of factors are involved. These factors can be summarised under two main headings: firstly, factors that arise from the measuring instruments themselves – TRW and ABC – and secondly, real changes in the relationship between the readership and circulation of a title.

(i) Differences between TRW and ABC

TRW and ABC are two independent systems of measurement, measuring two different things. ABC is an audit of the average number of copies of a publication that are sold (or in the case of free copies, handed to or picked up by customers) over a given time period. It is not possible to audit who actually reads a publication, so that is why TRW exists, to estimate the number and type of people who read particular newspapers and magazines.


ABC is regarded as a reliable measure of circulation. However, there are aspects of ABC reporting that can affect RPC. For example, in the case of sale or return, a number of the returns may not be completed in a particular period, so they appear in the next auditing period, and thus an ABC figure could understate the number of copies in circulation in one period and overstate the number in the following period, which could have the effect of increasing RPC in the former, and reducing it in the latter.

TRW – Sample Variation

TRW obtains its readership estimates by interviewing a representative sample of the population. TRW interviews some 3,500 people every month, but despite having a large, high-quality sample, TRW estimates, like any estimates obtained from a sample, are subject to sample variation. This is not a factor for ABC, which is not a survey but an audit. Sample variation is a function of the fact that samples are very rarely a perfect representation of the populations they are designed to represent, and although weighting can be used to correct the data, some variations between one period and another are unavoidable. The degree to which sample variation alone is responsible for period-on-period differences in the readership estimate for a given title can be demonstrated by what is known as significance testing. If the test shows the difference is significant, there is a strong possibility that it is a reflection of a real change in readership. If on the other hand the difference it is not significant, the difference may be due simply to sample variation and not to any real change in reading behaviour.

TRW – Methodological factors

TRW strives as far as possible for consistency over time in its methodology, in terms of the selection of the sample, the procedures used to weight the data, and the data collection method itself. However, the survey is also committed to innovation and improvement, and from time to time one-off changes may be introduced that could affect the level of readership claims, and therefore RPC, and in such cases it is possible that not all titles would be affected to the same extent.
Another consideration is the prompt screens (shown to respondents) to which titles are allocated. With any new titles coming on to the survey, and the closure of other titles, changes to these screens have to be made periodically, with some titles moving from one screen to another. It may be shown that when a title changes screen, this will not normally affect the level of readership claims, but in a small number of cases it can.

(ii) Real changes in the relationship between Readership and Circulation

In a past study of year-on-year changes in RPC across a number of titles on the GB NRS, around 35% of those changes were found to be statistically significant. In other words, in a third of cases the change in RPC was likely to be a reflection of a real change in the relationship between readership and circulation. There are a large number and variety of factors that can contribute to real changes in RPC, and a number of these are summarised below. These factors do not necessarily operate in isolation, but can be inter-connected, and one can counter another, so it is often difficult to identify and quantify with any certainty why the RPC of a particular newspaper or magazine changes over time.

a) Source of Copy

The way in which copies are obtained, and how they subsequently move about, is a key determinant of RPC. In principle, copies which are first read in-home, and stay there until they are disposed of, accrue fewer pass-on-readers (and therefore have a lower RPC) than copies that are passed on to other households, or those that are circulated in workplaces or displayed in other public places. Certain actions by publishers concerning the distribution of copies can therefore affect the level of RPC, for example promoting the circulation of copies in workplaces could in some cases lead to an increase in RPC.

b) Changes to the publication

Changes to the format or editorial content of a publication can affect its RPC. These changes can be slow and evolutionary, or, as in the case of a major relaunch, more dramatic. In the latter case, it is often associated with heavy promotion of the title. The effect on RPC can go in either direction: a major relaunch could be akin to the launch of a new title (see below) in which case readership not keep up with increases in circulation and RPC might therefore fall. On the other hand, changes to format or content could broaden the appeal of the publication in which case RPC could increase.

c) Short-term fluctuations

In addition to long-term planned changes by publishers, short-term fluctuations in circulation and readership can occur, and these can be of sufficient magnitude to be reflected in RPC. For example, major news items can increase circulation and readership of newspapers but if the two are not affected equally, RPC could go up or down temporarily. Similarly, the sales of certain magazines are heavily influenced by promotions such as cover-mounted gifts: an attractive promotion can substantially increase sales of a particular issue, but this does not automatically translate into an equivalent uplift in readership.

d) New titles

A newly-launched publication usually has a lower initial RPC than the figure it eventually achieves after 3-4 years. One reason is that a network of pass-on readership can take some time to be developed, in which case RPC of a new title starts low and slowly rises over time, and this would imply that the TRW estimate is a good indication of what is actually happening. A separate possibility is that because some new titles may take time to establish themselves in the public consciousness, a proportion of TRW respondents who have read such a title might not recall the title in the TRW interview, in which circumstances TRW could to some extent be under-estimating readership. Over time, as readers become more familiar with the title, the readership estimate – and RPC – could increase.

e) Competition from other titles

The launch of a new title can have an impact on the RPC of existing titles by affecting circulation or readership or both. For example, if pass-on readers of a title (i.e. those who do not buy or acquire it themselves) are lured away by the arrival of a new title, the circulation of the existing title might not necessarily be affected, but its readership, and consequently its RPC, could decline. Changes in editorial format and content of one title can also have an impact on the circulation or readership of its competitors. When a title closes, this can also affect other titles, insofar as readers of the closed publication migrate to other publications.

f) Increases in circulation tend to reduce RPC

When circulation of a title is growing, it is often the case that readership does not grow at a similar pace. As is the case for new titles, someone who has only recently started to regularly buy a publication will generally generate fewer pass-on readers than established purchasers. Moreover, if the growth in circulation comes from people who were themselves previously readers of someone else’s copy, not only does this mean that new readers are not being generated, but previous pass-on readers are being lost. An example of the latter might be a title whose circulation has been boosted by the success of a subscription programme: if a proportion of those taking up the subscription offer were existing readers, circulation would increase but readership would not, and therefore RPC would fall.

g) Free Copies

Copies of a publication that are free tend to generate fewer pass-on readers than copies that are paid for, because a person picking up or being handed a publication for free may feel less
inclined to hang on to it or pass it on than if they had paid for it. This may also be true, but to a lesser degree, in circumstances where copies are given away for substantially less than the full cover price.

h) Readership lag

In some cases, readership follows the same general trend as circulation, but lags behind it. Circulation of a title might increase, but readership stays level for a time. Conversely, the reverse might occur: circulation starts to fall, but again readership remains level for a time. Such lags can occur because publications can continue to generate readers long after their publication date; this is more likely to occur with magazines than with newspapers, and within magazines it is more likely to affect those titles with a long active life.

i) Multi-Platform development

Many publishers have invested in extensions of their printed
titles using different platforms, the aim being to increase the contact existing readers have with their publication brands, and to attract new customers. Websites and apps offer alternative means of accessing editorial content, and this could have an impact on circulation and readership of the print versions. Those who purchase the publication can use both – print and website/app – but if former pass-on readers of a print title now only use the website or app, this will result in a fall in RPC for the title concerned.

j) Household Size

Household size can have a strong influence on RPC, particularly for the types of publication that tend to stay in the home. Any long-term trends toward a greater number of smaller households and this could contribute to a long-term general decline in the RPC of newspapers and magazines.

k) Demographic Changes

Changes in the make-up of the population can affect the circulation and readership of some types of publication more than others. National population growth may not necessarily be evenly distributed by age. By way of illustration, one could see, for example, the number of 15-24 year-olds increase by 12%, and the number of 55-64 year-olds by 25%, whereas 25-34 year-olds decline by 16%. The increase in 15-24 year-olds has increased the potential pool of readers for youth-orientated titles, but this has to a degree been offset by an underlying decline in reading newspapers and magazines by this age group. The increase in the 55-64 year-olds has created new opportunities for publishers, but this age group tends to live in smaller households and in general has a smaller social circle than young adults, and therefore the potential for generating additional RPC is somewhat less. Titles with an ageing readership might be expected to show a decline in RPC.

l) Social Factors

In addition to demographic factors, there are some social changes that can affect the readership and circulation of some publications. For example, the increasing affluence of the population, and the tendency towards greater discretionary spending, could mean that people are in general more able and willing to buy a magazine or newspaper themselves rather than reading someone else’s copy, and this would have the effect of reducing RPC.


Readership and Circulation are two different things, and TRW and ABC are two different measures. If a change in one is not mirrored by a change in the other, it does not mean there is necessarily something wrong with either TRW or ABC. The relationship between Readership and Circulation is not fixed and the RPC of a title can legitimately vary across time.

Leave a comment