What is MMM (Marketing Mix Modeling)? Explaining its features, procedures, examples, etc.
Marketing Mix Modeling (Mmarketing Mix MMarketing Analytics (MMM) is a statistical analysis that quantifies the impact of marketing initiatives on results.
MMM is used mainly by companies with large promotional budgets in the United States, a country that is advanced in marketing, but its penetration rate in Japan has not been high. However, even in Japan, interest and adoption rates have been rising rapidly over the past 4 to 6 years.
The reason is that general "consumer research (quantitative and qualitative)""Tracking" and "log analysis" using cookiesMMM is highly regarded for its unique features, including its ability to comprehensively cover not only digital measures but also offline marketing measures such as TV commercials and OOH, while also considering consumer privacy and conducting analyses that would be impossible with conventional methods.
table of contents
What is MMM? What you need to know
Purpose of MMM
The main objectives of MMM are:Understand how various marketing initiatives drive business resultsIs to do.
How MMM works
At MMM, we break down the various factors related to business results,Distinguish between the contribution of advertising and sales promotion activities (incremental effect) and other (base factors).
Benefits of MMM
The main benefits of MMM are:Efficient allocation of marketing budgetand planningPredict results based on your marketing efforts(or another spending scenario you want to simulate).
Free download of case study
What is "MMM" that all marketers should know about?
~ Benefits of Implementation: Insights from Three Company Case Studies ~
The Global State of MMM: An External Survey by an Advertising Industry Association
Digging deeper into the statement that "in the United States, a country that is advanced in marketing, MMM is being adopted mainly by companies with large promotional budgets," we can find out the current state of MMM and the extent to which it is being used globally from a survey of 2019 companies conducted in 116 by the IAB (the non-profit organization "Interactive Advertising Bureau").
Global usage of MMM
A survey on marketing effectiveness measurement shows that over 66% of companies use MMM. Furthermore, nearly 8% of marketers use MMM. This means that MMM is used by the majority of companies globally.
Additionally, we can see that of the companies that use MMM, over 63% have introduced it more than five years ago and continue to use it as a method for analyzing marketing effectiveness.
*Source: The Interactive Advertising Bureau (IAB),The essential guide to MMM and MTA", n=116, 2019
Positions of MMM users and average ROI improvement
Looking at the job titles that utilize MMM, 8% are general managers and above, as well as executives and corporate officers. This shows that the more roles and responsibilities people have in decision-making to understand the overall picture of marketing effectiveness based on data and optimize ROI, the more essential MMM tends to become.
Also, in terms of benchmarking the results of your efforts, it is realistic to say that optimizing your budget allocation can improve your ROI by an average of 3-4%, especially in the case of continuous use.
*Source: The Interactive Advertising Bureau (IAB),The essential guide to MMM and MTA", n=116, 2019
Features of MMM
One of the features of MMM is that it can quantify the "impact on other marketing measures (hereafter referred to as the indirect effect)" and "impact on results (hereafter referred to as the direct effect)" of implementing a certain marketing measure.
In today's world, where media and channels continue to increase, marketers are required to ① use multiple media and channels and ② implement multiple marketing measures simultaneously. In order to maximize results in this environment,Optimization of each marketing measureIt is necessary not only to promote each marketing strategy, but also to maximize the synergistic effect of each marketing strategy. Since MMM can quantify the synergistic effect as an indirect effect, it is only natural that the attention and adoption rate will soar.
Another feature of MMM is that it can also include external factors that a company cannot control in its analysis.
Even if your company's marketing measures are perfect, if a competitor dramatically cuts their prices, sales will fall, and if a typhoon comes, the number of customers will decrease. Conversely, even if your company's marketing measures fail, some other factor may improve the results. With MMM, you can include such external factors in your analysis, and it is possible to quantify the pure effect of your marketing measures, excluding external factors.
Related services:MMM (Marketing Mix Modeling) Analysis Service
MMM Procedures and Methods
The procedure for MMM is
- Decide on the analysis logic
- Identify past marketing measures (= internal factors) and external factors that affect results
- Allocate the measures identified to each stage of consumer purchasing behavior
- Collect the data
- Do the analysis
- Improve analytical accuracy
It will be something like that.
1. Decide on the analysis logic
First, you need to decide (select) the analysis logic.
There is not just one specific analysis logic for MMM. The subsequent steps will change depending on which logic is adopted. In this article,MAGELLAN (hereinafter referred to as Magellan)The remaining steps are based on the analysis logic known as path analysis used in the previous section.
Path analysis is an analytical logic that defines the influence between variables as a linear relationship and predicts in advance the steps (causal relationships between variables) to reach the goal.
2. Identify past marketing initiatives (= internal factors) and external factors that affect results
Identify all marketing measures implemented in the past. What is particularly important here is to identify them at a granularity that matches the purpose of the analysis. For example, if you are running creatives with multiple appeals on a TV commercial and your goal is to visualize the effects of each appeal, you need to identify measures at the granularity of "TV commercials with appeal axis A" and "TV commercials with appeal axis B" rather than at the granularity of "TV commercials."
Next, we identify external factors that affect the results. This means factors that are beyond our control but have a significant impact on the results. For example, we cannot control whether a competitor's TV commercial will be broadcast, but if we think that it will affect the results of the product or service being analyzed, we identify "competitor's TV commercial" as an external factor.
3. Allocate the identified measures to each stage of consumer purchasing behavior
Once you have identified your marketing strategies, define consumer purchasing behavior. See Image 1 for an example of the output.
Taking AIDMA, a well-known model of consumer purchasing behavior, as an example, consumers become aware of a product or service (Attention), become interested in it (Interest), want it (Desire), remember the product or service (Memory), and then make a purchase (Action).
Consumer purchasing behavior varies depending on the product or service, so please decide on a model that is appropriate for the product or service being analyzed, even if it is based on a hypothesis.
After that, the marketing measures that are considered to be most effective for consumers are assigned to each stage of consumer behavior (in the case of AIDMA, awareness, interest, desire, memory). For example, if there is a hypothesis that consumers will perform name searches for products/services after watching a TV commercial (increasing the volume of name searches), then TV commercials should be assigned to a step before name searches on listings.
4. Collect data
Next, collect data on the measures at the granularity identified in "(2) Identify external factors that may be affecting past marketing measures and their results."
We collect data on each marketing initiative (advertising volume and cost) and results data (sales numbers, costs, etc.) on a daily or weekly basis.
5. Conduct analysis
We will conduct an analysis based on the logic decided in "(1) Decide the analysis logic".
At this point, there are some points that need to be considered that are specific to marketing analysis, so you need to be careful.
6. Improve analytical accuracy
Generally, the accuracy of the initial analysis will be low, so various adjustments will be necessary to improve the accuracy of the analysis. For example, if the step to which a certain marketing measure should be assigned is different from the one assumed in the stage of "(3) Allocating the measures identified to each stage of consumer purchasing behavior," assign it to a different step and run the analysis again to check whether the accuracy has improved.
Furthermore, if it is found that accuracy has decreased due to unexpected external factors, those external factors must be included in the analysis, so data on those external factors must be collected and the analysis must be performed again to see whether accuracy has improved.
MMM Case Study
There are not many published examples of MMM, in fact they are almost non-existent, because MMM reveals facts that are highly relevant to the marketing of a product or service.
MMM is implemented with the goal of optimizing marketing overall, not optimizing each marketing measure. Therefore, much of the information revealed by MMM is things that you would not want your competitors to know, such as "the synergistic effect of implementing measures A and B simultaneously is very large," "the influence of external factor A is very large, and it has a particularly large negative effect on measure C," and "the most important KPI for increasing sales is indicator D," and there are major hurdles to disclosing the results of actions based on that information.
Our MMM Analysis Solution"MAGELLAN"So, without changing the promotion budget itself,Optimize budget allocation for each marketing initiativeOn average, results have shown that just doing this improves performance by several percent.
Meanwhile, our clients explain why they decided to introduce MMM in the MAGELLAN case study interview, so be sure to take a look.
MMM (Marketing Mix Modeling) Solution Introduction Interview
Things to keep in mind when using MMM
When utilizing MMM, it is important to keep in mind that its purpose is different from general "consumer behavior research (quantitative and qualitative)" and "log analysis."
As a typical example, we will explain the difference between consumer research (quantitative) and MMM.
Consumer surveys (quantitative) are surveys that can reveal changes in consumer purchasing psychology due to advertising, etc. For example, by conducting consumer surveys before and after the broadcast of a TV commercial, changes in "awareness," "favorability," "first recall," and "purchase intention" for a company's products and services can be clarified. Consumer surveys are suitable for confirming whether the intention behind the advertisement was accurately conveyed and whether the intended response (change in purchasing psychology and attitude) was obtained. However, knowing the extent to which purchasing psychology changed does not tell us whether or to what extent the results, such as sales volume and number of contracts, actually increased.
While MMM cannot comment on changes in purchasing psychology, it can reveal the indirect and direct effects mentioned above, making it possible to make statements such as "If the viewership rate of a TV commercial increases by 1%, sales will increase by 10."
Unless you fully understand and are aware of what MMM can and cannot reveal,Improve your marketing and promotionsThere are various types of surveys and analyses in the marketing field, each with different objectives, so if your budget allows, you should conduct each of them, but if you have a limited budget, you need to choose the analysis that is most appropriate for solving your company's current problems.
Free download of case study
Can the effect of branding advertising on sales be proven?
~MMM reveals whether you should continue running TV commercials~
Summary of MMM (Marketing Mix Modeling)
MMM refers to statistical analysis that quantifies the impact that marketing initiatives have on results. It is being increasingly adopted in the United States, a market-leading country, and is also gaining attention and adoption in Japan.
The distinctive feature of MMM is that it can clarify the direct effect of each marketing measure on results and its indirect effect on other marketing measures, and it can also include external factors in the analysis.
By accurately understanding the characteristics of MMM and being aware of how it differs from other research and analysis methods, you will be able to improve your marketing.
If you want to learn more about MMM, check out our comprehensive guide to what marketers need to understand and pay attention to in order to make their MMM projects successful and lead to real action and contributions to their companies.MMM Practical Guide
↓ Click here for a list of carefully selected articles to deepen your understanding of MMM
- What is MMM (Marketing Mix Modeling)? Explaining its features, procedures, examples, etc.
- Why you should use Marketing Mix Modeling (MMM) to measure the effectiveness of your advertising
- 10 Common Misconceptions About Marketing Mix Modeling (MMM)
- What are the limitations and areas in which MMM (Marketing Mix Modeling) does not excel?
- Latest Trends and Future Outlook of MMM (Marketing Mix Modeling) (2024 Edition)）
- “Modern MMM (Marketing Mix Modeling)” Practical Guide for Marketers
