Australia’s mapp makes better estimates of return-on-investment
An Australian initiative marks a step forward in assessing the cost-effectiveness of advertising in magazines: the Magazine Audience Performance Predictor, or ‘mapp’ for short. It involves forecasting how advertising messages in magazines will be spread through time, issue by issue.
In addition to using readership accumulation curves (a necessary but often neglected step in assessing magazine campaigns), mapp uses weekly scan data of magazine sales to predict the final total audience of each printed issue of each magazine brand. This is then built into a campaign planning tool which publishers and media agencies can use.
The greater precision promised by this development should mean that the link between magazine advertising and advertisers’ product sales will be measured more efficiently, resulting in more accurate – and higher – estimates of magazines’ return on investment.
Mapp has been developed by Magazine Publishers of Australia (MPA), with much of the work conducted by MPA member publishers.
Problem 1: measuring campaigns’ weekly delivery of ad exposures
The weekly delivery of ad exposures is a vital aspect of proving magazines’ cost-effectiveness, because of the way that the econometric models work which are used by advertisers and their agencies in assessing their media choices.
Most commonly, weekly product sales data are compared with weekly data on a host of other variables, including consumers’ exposure to each advertising medium used. The models work out how closely each medium’s weekly fluctuations in audience delivery match the weekly fluctuations in product sales. The closer the match, the stronger the assumption of a causal effect – represented by the calculated return on investment (ROI).
For magazines, with their audience build-up spread across time (not just weeks, but months, for an individual issue), the easy assumption for modellers is that the entire audience to an issue of a magazine occurs in the week of publication; or that the audience is spread evenly across all weeks till the next issue. This means that the link between exposure to magazine ads and weekly sales of the product is shown to be much weaker than it really is, thus seriously underestimating the medium’s ROI. Or, as MPA put it, “a large share of magazines’ performance is lost as models look for inputs in the wrong places”.
That is why it is vital to use readership accumulation curves. They show how the total readership of an issue is spread through time, and thus provide a much closer and more accurate link between ad exposure and product sales.
Problem 2: variations in issue by issue readership
There is another problem. The usual measure of the readership of a single issue of a magazine is that of an average issue; no account is taken of variations from one issue to the next, which can be significant. In a small minority of cases fluctuations in copy sales are taken into account, using the hindsight sales of specific issues. But these data are not available until long after the event, restricting their value for current planning.
Mapp’s solution: current specific-issue audiences
The innovative feature which mapp has introduced is to predict current issue-by-issue audiences, magazine by magazine. That is, a forecast of the audience each issue would achieve, calculated shortly after publication of the issue, and before much of this reading had actually taken place.
Crucial raw data is scan-based sales data covering the first week of publication of each issue. A formula has been devised which has proved very successful in projecting the first-week data to create a good estimate of final readership numbers.
The mapp tool
Each Monday the computer services bureau Quantium uses the latest weekly scanned sales data and other inputs to calculate current audience estimates for current and recent specific issues of all 31 magazines in the system. Known as the mapp tool, this makes it possible to analyse week by week audiences (as gross rating points) delivered by any campaign using any of these 31 titles.
Very importantly, this is a stand-alone tool which can export data in the required format for agencies and advertisers to use in their own econometric or other models.
It is too soon after launch of mapp for any case histories to be available yet, but some are in progress. However it seems highly likely that the case for magazine media will be much strengthened in Australia.
Discussion at Hamburg in June
Mapp will be one of the topics discussed at the FIPP Research Forum in Hamburg on June 16-17.
No doubt delegates will be interested to discuss how well the predicted single-issue audiences match the audiences reported in the industry readership surveys; what difference the specific-issue data make compared with normal practice; the take-up of the mapp tool by media agencies; ROI comparisons between magazines and competing major media; other forms of case study; further refinements of the system; and how far delegates can or should apply the learnings from mapp to their own countries.
If you have any comments on this article please email FIPP’s Research Consultant, Guy Consterdine