Ipsos, Outdoor Special Power (OSP) methodological changes and Reach
Methodological changes in Outdoor Audience survey
During the publishing of the 2011 2nd half database of the outdoor audience measurement, the planning software itself has been renewed. The databases in the new Outdoor Special Power are based on an updated methodology, which can provide more accurate results.
These results can be lower than with the previous methodology. But with thorough investigation it can be seen that the results did not drop significantly for all of the panels, it only dropped for the lower quality panels, thus bringing the results of the whole campaign closer to reality. The previous methodology was too “permissive” (differences between campaigns did not came out clearly in the results) so it was necessary to apply a new methodology.
The basic difference between the old and new methodology is the territoriality. While the old one searched basic contacts in the whole visibility zone of the panels, the new searches only on those route elements, which are assigned to the panels (of course within the zone of visibility). The main difference is that only one route element (the base unit of the survey) can be connected to each direction of a panel. The other difference is in the modelling phase, where the modelled contacts were reduced thus reaching the level that is closer to reality.
The new methodology result more accurate Reach results and these results are closer to reality. These results were controlled with results from POSTerTEST campaign effectiveness measurement.
The Reach indices
Simple reach
When counting Simple Reach we compute all the contacts of each panel in the campaign with respondents in the database. Every respondent gets an aggregated contact value, depending on how many panels did he or she met. In case of Simple Reach those respondents will be counted as reached, who got at least one contact with any of the panels in the campaign.
It can be seen on the diagram that after the first contact (GI), the person is counted as reached.
Effective reach
The computing of Effective Reach is based on the same methodology as Simple Reach, the only difference is that respondents are counted as reached not after one but after they get at least N contacts. This N parameter can be set for each campaign, although in the Hungarian market 8 is used. This Reach index with the use of parameter 8 is much closer to results from POSTerTEST campaign effectiveness measurement but it still can be fine-tuned.
It can be seen on the diagram that after the eighth contact the person is counted as reached. The logic behind this shows that when someone sees the campaign only 7 times (has 7 contacts) he or she is not reached by it, but when that person gets one more contact then he or she is reached.
Memorisation Reach
The Memorisation Reach is the most accurate index. The counting logic is different from the previous two Reach indices. This index does not “cut” so roughly the reach, in other words there is no strict limit which a respondent needs to pass to be counted as reached – this limit is the contact number.
Every respondent gets a personalized Reach value calculated (with the given Memorisation Reach parameter), which can be between 0 and 1. This way every respondent gets a reach chance, and the Reach does not depend on whether the respondent met the campaign once or 8 times.
The function of Memorisation Reach is: 1-(1-p)ˆk, where ’p’ is the memorisation parameter, and the ‘k’ is the contact number.
The parameter of Memorisation Reach is 0,13.
It can be seen on the diagram that as the contact numbers grow, so grows the chance of being reached. This way with only 1 contact the respondent has a chance to be reached, and even with 30 contacts it is not automatic that he or she is reached.
This is the index that’s the most adequate to calculate Reach, for it gives the chance for everyone to be reached, while the results aren’t unrealistically high as in case of Simple Reach.