Cross-Channel Marketing Mix Modeling, or Media Mix Modeling, is the channel analytics process that utilizes historical info to determine the value of varying marketing strategies on the impact of sales. Historical information like a corporation’s internal data and third-party point of sale information is reviewed and analyzed in this effort. The end result is to produce a quantifiable and comparable relationship between the different marketing endeavors to show which media channels are more useful and productive for the sales of a given product or various products. These relationships are quantified by their effectiveness of sales volume that is generated, by the volume produced per unit of input, by the efficiency in terms of cost divided into sales volume, and by the return on investment. The final product uses predictive analytics to produce a list of recommendations on which marketing strategies will be most productive so that the actual marketing and promotion can be adjusted and finally optimized.
Where Did Marketing Mix Modeling (MMM) Come from and When Did it Achieve Marketplace Dominance?
In light of how successful Marketing Mix Modeling, or MMM, has become, it is hard to imagine that it is a fairly new field of study. Only during the past decade did a number of Consumer Packaged Goods companies fully accept MMM. Nowadays, a great number of the important Fortune 500 firms including Pepsi, Coca-Cola, Kraft, and Proctor & Gamble have integrated MMM into the heart of their marketing strategies. The success of MMM in achieving world dominance stems in part from the availability of well-regarded marketing firms that specialize in the MMM services.
From the CPG segment, marketing mix models migrated successfully into Pharmaceutical and Retail businesses. These were natural fits with MMM because of the third party data that is available to them through Symphony IRI Group, Nielsen Company, and the NPD Group. This time sensitive information is necessary for MMM efforts to be successful. Thanks to CRM systems customer data availability, auto, hospitality, and financial services industries also picked up the MMM strategy in recent years. Financial services companies were able to rely on the third party data provided by Forrester Research’s Ultimate Consumer Panel in order to make efficient MMM possible in their field.
What Were the Problems Financial Services Companies Faced before Using Marketing Mix Modeling?
Numerous problems plagued the financial services industry’s marketing and advertising efforts before the advent of Marketing Mix Modeling. There were disparate data sources that could not always be quantified or objectively compared against one another. The metrics to measure one marketing effort against another were often totally incompatible and not translatable. Financial companies were desperately in need of some effective form of cross-channel media mix modeling. They were attempting to optimize their marketing media mix from a misinformed understanding of what strategies were working and what were not. They also did not have any intelligence from their end user client audience. Further, these companies could not at all predict the results of any given campaign effort. All of these complex problems ultimately led to a sub-optimal mix of marketing channels.
How Do Marketing Mix Models Solve These Problems?
Market mix models solve these problems effectively and efficiently with several inventive solutions. It breaks down all channel analytics and determines their contribution on a channel by channel basis. Next, it crunches the costs associated with these channels to determine their individual activity’s Return on Investment. MMM evaluates the marketing activity’s overall effectiveness. With all of this information in hand, it is able to engage in predictive analytics in order to ensure that the marketing spending distribution is optimal.
How Does Marketing Mix Modeling Specifically Support Financial Services Companies?
Such MMM activities have specific helpful applications for Financial Service Companies. They provide bankers with a superior capital and credit risk management strategy, better operating efficiency, higher profit margins, and stronger relationships with customers. They retool the various promotions to reflect the most appropriate products for interest rate discounts. In financial services, these are the main promotional tools in this highly competitive industry.
How Does Marketing Mix Modeling work?
The main function of Market Mix Modeling is to translate the value of marketing efforts into a direct and demonstrable connection to something happening in sales, market share, and return on investment. To do so, the practice focuses on comparing aggregate historical data between marketing and sales metrics.
- The MMM approach is based on a popular marketing theory known as the 4Ps of the Marketing Mix, which are: product, price, place, and promotion. This theory states that these are the four fundamental elements of any successful business. The purpose of MMM, then, is to measure how much success came from each of these elements, and create projections for maximizing success through optimization of the mix.
- In order to do so, MMM compares aggregate data to understand, identify, and differentiate which factors led to the specific success. Was it marketing efforts and promotions? Or was there some external factor at play?
Throughout the process, marketing mix examples rely on the use of multi-linear regression to identify a correlation between an independent variable x and a dependent variable y, where the value of y can be predicted by measuring x.
Dependent variables represent the hard, financial data that illustrates success. This might mean sales, market share, or stock price.
The independent variable represents the marketing efforts. These are broken into two categories; Above the line (ATL) and Below the line (BTL). ATL activities would include print, radio, TV, and digital ads. BTL activities are temporary promotions, sales, coupons, contests, and direct mail marketing.
- From here, marketers form an equation between two variables. The line drawn from this equation could be linear or nonlinear, i.e. connected or not connected.
- For example, a TV or digital advertisement may have an effect on brand awareness, but it can’t have a linear relationship to an increase in sales; rather, we need to look to non-marketing factors. These are known as base drivers, and can include price, distribution trends, seasonality factors, and other macroeconomic influences.
The Benefits of Marketing Mix Modeling
By identifying and measuring the discrete factors that led to a specific instance of success, marketers can draw educated and informed conclusions. This makes it much easier to create blueprints for future growth, specifically by:
- Analyzing marketing spend and budgets to get the most out of every dollar. Consider the non-linear relationship between advertising and sales; most models show that while ad spend can increase awareness, there’s a limit to its efficacy. After a certain point, spending any more on ads could be a total waste of money; in another scenario, the push from ads may not come until after a certain dollar threshold. MMM helps to establish those price points
- Establish accurate forecasting for future marketing strategies. With a handle on which marketing efforts work best, as well as which ones are trending up (or down) teams can plan out their strategies further into the future.
- Uncovers hidden or ignored correlations that the company could lean into. For example, it may be hard to explain, but maybe people in Minnesota just like a car more than people in Detroit. You can’t change instinct, so why not create more marketing around those correlations?
In general, MMM provides high-level insights into marketing campaigns and strategy, as well as shedding light on the trends that could be most impactful.
How to Use Marketing Mixed Modeling
While these techniques are mathematically sound, there have also been claims that the MMM is a kind of dead marketing language, done away by time and more advanced tech tools. For instance, there’s the concern that MMM doesn’t provide enough insights on the consumer level, or help marketers to create customized messaging. Additionally, the use of historical data over the course of two or three years means infrequent reporting.
However, MMM can be greatly beneficial when it’s performed once or twice a year as part of a larger marketing strategy. This allows marketers to still benefit from those high-level insights, and then keep those in mind when using more granular data analysis techniques like data-driven or multi-touch attribution. MMM also provides insights into the kind of offline conversions that occur in-person with a sales team.
