New MMM insight from JCDecaux UK

New MMM insight from JCDecaux UK

1st May 2024

An important new insight from JCDecaux UK is a call-to-arms to change the way that Marketing Mix Modelling (MMM) is conducted, showing for the first time the crucial role that postcode-level data plays in proving the ROI of media channels including Out-of-Home (OOH). The research conducted by Nielsen and commissioned by JCDecaux shows that OOH’s contribution is being undervalued in current econometric models and, in our research, adding granular data increased the return-on-investment (ROI) of OOH by 42%, showing OOH’s true impact on sales in the models for the first time.

The findings were unveiled at a Research seminar ‘Location Matters for MMM’, held in London on 30 April to an audience of Research and Measurement experts.

The ‘Location Matters for MMM’ insight was conducted by Nielsen across 3 retail brands in 2021/2022 and shows that by including location-based impressions and sales data (Route & CACI), OOH’s ROI rises by 42%.

Key findings from the research:

  • Does adding granular data provide more accurate ROI for media channels in MMM?

Result: Yes. As an example, OOH’s ROI increased by 42%.

  • Does adding granular data increase the correlation between sales data and media channels?

Result: Yes. OOH’s correlation shows an increase of 78%. From a correlation of sales-to-transactions of 45% in current MMM to an 80% correlation in MMM using granular data - explaining why ROI for OOH increases.

  • Does adding granular data change the overall accuracy of the model?

Result: No. There is no change. Using a single value to assess the whole model and therefore accuracy of the output for individual channels can be misleading.

Chris Felton, Director of Data and Insight at JCDecaux UK said, “MMM econometric modelling is vitally important as it attributes sales to marketing activity, enabling brands to make investment decisions because they can see which channels delivered best for them. From an Out-of-Home perspective it’s critical that our true sales effect is captured by MMM. We are calling for marketers, agencies and researchers to ensure their MMM uses the most accurate data possible both in terms of the granularity of location and time. This will help identify what's really driving their sales, as a location based medium our impact will also vary by geography and should be measured accordingly.”

Jeremy Cartwright, Consulting Director EMEA from Nielsen said, “The ‘Location Matters for MMM’, study shows that including postcode level data in MMM models provides an unprecedented view of what each channel is delivering in terms of sales, but importantly it benefits the accuracy of how they measure.  For example, the movement of people is key to accurately capture how potential consumers interact with OOH advertising. “

The ‘Location Matters for MMM’, work conducted by Nielsen and commissioned by JCDecaux follows a study by Outsmart & Entropy  that outlined the three main limitations of current MMM in terms of capturing Out-of-Home’s contribution. This earlier study showed:

  • National data: MMM use national aggregated figures which don’t capture OOH’s impact on sales in specific areas. 62% of MMM practitioners said this is a barrier to OOH measurement.
  • Collinearity: MMM is unable to separate out the effects of channels running at the same time or the effect of priming other channels in terms of attributing sales.  38% of respondents said this is a barrier to OOH measurement.

Size of execution: As a 5% medium, Out-of-Home is usually booked as part of a larger campaign. 55% of respondents cited campaign investment being too small as a barrier to measurement.