How a leading consumer packaged-good company evolved their marketing mix modelling outputs and media planning strategies with Arima
A leading Global CPG marketer with multiple well-established brands under its umbrella partnered with Arima to better understand the impact of their marketing activity on in-store sales. To deliver these insights Arima deployed it’s high-speed Marketing Mix Model platform.
MMM Objectives:
- Analyze historical media spend to uncover which channels were driving the best return on ad spend (ROAS)
- Understand which media channels could deliver incremental sales if media dollars were re-allocated
- Establish an optimal media mix for each brand in the advertiser’s portfolio by total market, region and also by target audience
- Define the sales impact of competitive brand activity, in- store promotions and related macroeconomic factors
The Approach & Result
Arima analyzed two years of sales data and media spend across both online and offline media channels to establish a correlation between all in-market media activity and revenue.
Return on Ad Spend (ROAS)
ROAS was allocated and reported by channel. In the chart below two important metrics are provided; historical ROAS (orange bars) and marginal ROAS (blue line).
ROAS looks backwards and shows how much revenue was generated for each dollar invested in the channel.
Marginal ROAS looks forward and defines how much revenue each additional dollar spent in the channel would generate. Please note, marginal ROAS details include a declining return on investment curve (not shown) as media become less effective as more budget is allocated to that channel.
As performance of each media channel is reported, the marketer can see that
search marketing (SEM) and Pinterest had the highest ROAS
Waterfall of Contributing Factors Affecting Sales
Marketing mix models evaluate both internal and external factors that contribute to a brand’s sales. The Arima platform provides flexibility to ensure significant market factors that affect your brand are taken into consideration.
For the marketer in this case study, factors incorporated in the Marketing Mix Model included:
● Competitive activity
● Major market events (lockdowns etc.)
● In-store promotions
● Sister brand activity
● Paid and owned media channels
● Base brand sales; what sales would be if the brand did nothing
The model shows the contribution of each marketing activity in-market and reveals a degree of cannibalization between sister brands
Forecasting Sales with a Live MMM
Within the platform the CPG marketer could also view sales forecasts for the upcoming quarters to conduct multiple ‘what if’ performance scenarios by adjusting spend allocations by media channel.
This easy to use sales forecast tool allows marketers to use learnings from previous analysis to explore multiple spend scenarios before investing.
As shown, SEM and Pinterest spend have been increased based on insights
derived from the related ROAS and Waterfall views of the data (prior tables)
Making MMM Forecasts Actionable in Media Planning
Finally, we used Arima’s audience builder function to define a segment “Consumed product past 30 days – YES,” along with media allocations suggested by the MMM optimizer. Within this adjacent tool the marketer could view and compare detailed KPI projections including:
● Campaign Impressions
● Campaign Reach and Frequency
● Total Campaign CPM
● Sales
The outcome analysis is not based on the advertiser’s spend, rather it compares two fictional campaign scenarios with the same media spend and audience but varying media allocations.
The output of the media planning tool suggests that an increase in sales of 2.4%
can be achieved with a channel reallocation. However, this reallocation will result
in reduced impressions, reach and campaign frequency and a higher campaign
CPM. This helps the advertiser manage multiple KPIs while keeping a focus on
primary vs secondary metrics.
About Arima
Arima is North America's most advanced marketing platform, combining live market mix modeling, consumer insights discovery, and media planning tools into one source.
The Synthetic Society by Arima enables a ‘privacy by design’ data fusion for marketers and agencies. Our suite of tools allows hundreds of organizations to unlock the power of marketing data science in easier and more affordable ways.
Contact us or book a demo to learn more.