Haik-OOH - Location Matters
Location matters
Add audiences from Route
Make profit increase
The requirement for accountability in ad spend has never been higher and with this we see the continued rise and rise of marketing mix modelling to help demonstrate return on investment. As with any data model, the outputs are only as good as the inputs and, historically, OOH has not fared well in these sorts of models. This is more a reflection of the data provided to represent OOH rather than a commentary on the medium as a whole.
Recent work by JCDecaux, Talon and Nielsen has aimed to break the status quo. This piece shows the demonstrable effect on the performance of OOH within a MMM when highly granular (geographically and temporally) audience data from Route is used as an input in place of a more generic ‘national spend’ figure.
The novel methodology deployed in the work added geographic catchment areas from Route into the model inputs, in doing this the ROI for OOH increased by 42% than a version not including this level of data and increased the correlation with sales by 78%.
You can see more information about their work on the JCDecaux website here.
The granularity of OOH audience data available through Route is significant. Audiences (impacts) can be output for each poster or screen for every 15 minutes of the day, for each day of the week, for each month of the year.
Get in touch if you want to find out more about Route audiences and how they could help your data model better reflect the true effect of OOH.
Appraising MMM
As part of a cross industry initiative pioneered by Outsmart, the OOH industry came together with the ROI Fightback. Among other things, this identified key questions to ask when undertaking media mix midelling. These include:
Did the creative execution match the ROI metric?
Was there variation in the OOH data, or was it allocated equally across two week period?
Was an ROI score produced for each channel?
Was OOH included in the final model?
Was Route data used as the OOH measurement?
Was the OOH spend high enough to be picked up in the model?
If the answer to any of these is “no” then there could be some challenging results ahead.
How to represent OOH in MMM?
See here for more information on how Route data can be used in Econometric / Media Mix Modelling.
