A leading car manufacturer initiates efforts to develop a media mix model that will maximize marketing ROI. Channels include TV, out-of-home, digital (video, brand-paid search, non-brand paid search, social, affiliate, etc. ). TV has both national and local ads (hence local and national spending).
The marketing spends data for each channel is weekly for non-digital channels. The conversion label refers to the total units of sales for each of the three brands.
If you check any critical business questions below, Marketing Mix Models (MMM) could be of help
How much of your sales can be attributed to marketing efforts?
What is the contribution of TV, radio, and digital channels towards your business KPIs?
How do marketing channels interact and what are the trade-offs between channels?
What is the network (across channels) or halo (across brands) effect?
How effective are each of the marketing channels in impacting sales?
Marketing Mix Models (MMM) can help businesses to maximize the ROI across all marketing channels and optimize future spending for all media channels.
Data Requirements
At least 2 years of weekly spend, units, and pricing data
Data aggregated at location, product and time
Marketing communications and engagements
External market conditions
Data Types
Incremental drivers: Targeted business communication designed to increase conversions such as TV and print ads, digital media, price discounts, promotions, social outreach, etc. These are controllable variables.
Base drivers: General business communication to determine conversions without advertisements such as brand equity, general pricing strategies, and membership programs. These are both controllable and non-controllable variables.
Other drivers: Factors outside of the control of the client such as weather, macroeconomic policies, industry conditions. These are non-controllable variables.
Customized Model Methodology and Algorithms
Algorithm transformations include diminishing returns estimations, ad stock, and Bayesian prior results.
Marketing Mix Model (MMM) or Multi-Touch Attribution (MTA) Model?
A Marketing Mix Model (MMM) or Media Mix Model tells you how effective are each of the marketing channels in impacting sales. The marketing channels usually include TV, Radio, Digital Channels, Direct Mails.
Multi-Touch Attribution (MTA) Model tells you how effective is each touchpoint of a specific marketing channel in impacting sales (channel conversion rate). These touchpoints are at the most granular level.
FocusKPI team combines MMM and MTA models to design customized solutions that tailor your business needs. We’ve invested extensively in a suite of tools and technology that enable us to extract the meaningful connections, correlations, and trends from your behavioral and transaction data. Our expertise in these disciplines enables us to apply theories and inventions to create a breakthrough for your marketing campaigns.
In this case study, over a year, the FocusKPI model helped the client gain a 15% percent increase in marketing ROI over the three brands.
Data for the model was weekly spending and sales data for each geographic region. Separate regional algorithms were “fit” to the data. Results showed increasing marketing investments yield diminishing returns.
External macroeconomics variables included the S&P index, competitor pricing, seasonality, and holidays. The relationship between marketing spend and the units sold is estimated by the model based on historical data. Predictions from the model for one of the brands, with an average prediction error of 4%.
The attribution of each channel on units sold is calculated based on the model using proprietary algorithms. The optimal allocation of the planned marketing budget is based on the contribution of each channel and be different than the historical spend, in order to maximize ROI.
A leading car manufacturer initiates efforts to develop a media mix model that will maximize marketing ROI. Channels include TV, out-of-home, digital (video, brand-paid search, non-brand paid search, social, affiliate, etc. ). TV has both national and local ads (hence local and national spending).
The marketing spends data for each channel is weekly for non-digital channels. The conversion label refers to the total units of sales for each of the three brands.
If you check any critical business questions below, Marketing Mix Models (MMM) could be of help
How much of your sales can be attributed to marketing efforts?
What is the contribution of TV, radio, and digital channels towards your business KPIs?
How do marketing channels interact and what are the trade-offs between channels?
What is the network (across channels) or halo (across brands) effect?
How effective are each of the marketing channels in impacting sales?
Marketing Mix Models (MMM) can help businesses to maximize the ROI across all marketing channels and optimize future spending for all media channels.
Data Requirements
At least 2 years of weekly spend, units, and pricing data
Data aggregated at the location, product, and time
Marketing communications and engagements
External market conditions
Data Types
Incremental drivers: Targeted business communication designed to increase conversions such as TV and print ads, digital media, price discounts, promotions, social outreach, etc. These are controllable variables.
Base drivers: General business communication to determine conversions without advertisements such as brand equity, general pricing strategies, and membership programs. These are both controllable and non-controllable variables.
Other drivers: Factors outside of the control of the client such as weather, macroeconomic policies, industry conditions. These are non-controllable variables.
Customized Model Methodology and Algorithms
Algorithm transformations include diminishing returns estimations, ad stock, and Bayesian prior results.
Marketing Mix Model (MMM) or Multi-Touch Attribution (MTA) Model?
A Marketing Mix Model (MMM) or Media Mix Model tells you how effective are each of the marketing channels in impacting sales. The marketing channels usually include TV, Radio, Digital Channels, Direct Mails.
Multi-Touch Attribution (MTA) Model tells you how effective is each touchpoint of a specific marketing channel in impacting sales (channel conversion rate). These touchpoints are at the most granular level.
FocusKPI team combines MMM and MTA models to design customized solutions that tailor your business needs. We’ve invested extensively in a suite of tools and technology that enable us to extract the meaningful connections, correlations, and trends from your behavioral and transaction data. Our expertise in these disciplines enables us to apply theories and inventions to create a breakthrough for your marketing campaigns.
In this case study, over a year, the FocusKPI model helped the client gain a 15% percent increase in marketing ROI over the three brands.
Data for the model was weekly spending and sales data for each geographic region. Separate regional algorithms were “fit” to the data. Results showed increasing marketing investments yield diminishing returns.
External macroeconomics variables included the S&P index, competitor pricing, seasonality, and holidays. The relationship between marketing spend and the units sold is estimated by the model based on historical data. Predictions from the model for one of the brands, with an average prediction error of 4%.
The attribution of each channel on units sold is calculated based on the model using proprietary algorithms. The optimal allocation of the planned marketing budget is based on the contribution of each channel and be different than the historical spend, in order to maximize ROI.