10 Common Misconceptions About Marketing Mix Modeling (MMM)

2024.10.17 Column

As data-driven decision-making becomes increasingly important in the marketing field, an analytical method known as "MMM (Marketing Mix Modeling)" is attracting attention.

However, despite its popularity, there are also negative impressions of this method, such as "MMM is expensive and only large companies can use it" and "MMM is too theoretical and not practical."

In this article, we will address some common misconceptions about MMM and clarify the truth behind them. We hope that by clearing up these misconceptions, you will be able to understand the true value and potential of MMM and use this information to develop more effective marketing strategies.

table of contents

Misconceptions about usability and usefulness
Misconceptions about data and accuracy
Functionality Misconceptions
Why do these misconceptions arise?
Eliminate misconceptions and recognize the diversity of MMM
In conclusion

Misconceptions about usability and usefulness

1. Only large companies can use MMM effectively

Many marketers who are familiar with MMM assume that it is only beneficial for large companies with large advertising budgets. This is because in the past, MMM was primarily adopted by large companies. However, MMM can be scaled to a variety of scales and can provide valuable insights for businesses of all sizes.

Regardless of the size of your company, it is a good idea to consider the following points when deciding whether or not you can conduct an MMM analysis.

Quality and quantity of data:MMM requires a sufficient amount of high-quality data, typically at least two years' worth of detailed marketing and sales data is ideal.
Diversity of marketing strategies:The more a company uses multiple marketing channels and mediums, the more insights it can gain from MMM.
Organizational analytics resources:It is important to have staff in-house with the skills in statistics and data analysis, and to have a system in place for collaborating with external experts.
Budget flexibility:You need the flexibility to adjust your marketing budget based on MMM analysis results.
Executive Support:Management support that values data-driven decision making is crucial for the implementation and practical application of MMM.

2. MMM is not practical

Many marketers may think that MMM is too theoretical to be practical. However, when used correctly, MMM can provide useful suggestions to guide marketing strategies and budget allocation. For example, at XICA, many clients who provide MMM services use the suggestions to optimize the allocation of advertising budgets and improve ROI and sales. In this way, MMM is not just a theory, but a method that is directly linked to actual business results.

3. MMM is a one-off effort

Some people think that you can build an MMM analysis model once and use it forever. However, in reality, the analysis model is the core of MMM and should be updated continuously and flexibly to adapt to market changes and marketing strategies. Updating and adjusting it monthly or quarterly as needed will enable you to rotate the PDCA cycle quickly.

4. MMM is too complicated to implement

Due to its complexity, MMM tends to be avoided. It certainly requires a lot of data and specialized knowledge, but advances in technology and software are making MMM more accessible and easier to implement. For example, XICA's MMM service "MAGELLAN" uses proprietary technology to systematize the analysis process and trial and error, and also provides support from a team of consultants and analysts with specialized knowledge to help client companies quickly implement MMM, creating an environment that makes it easy to use MMM.

Misconceptions about data and accuracy

5. MMM needs perfect data

Some people believe that data used in MMM should not be missing, but in fact MMM can work even with incomplete data and provide sufficient insight. Of course, the more complete the data, the more accurate the model will be, but even if you do not have all the daily data or detailed variable data, you can take measures such as using alternative indicators, and you can take a flexible approach according to the situation. The important thing is to first formulate a hypothesis by consulting with your analysis team and MMM experts about your business issues and goals before pursuing the completeness of the data. Next, consider the data needed to verify the hypothesis and be flexible.

6. MMM's analysis results are always accurate

One might expect MMM to always provide accurate answers. However, MMM's analysis results are based on statistical models and are subject to a certain degree of uncertainty. These results should be used as guidelines, not as absolute truths. The key is to use the analysis results to test hypotheses and make continuous improvements.

7. MMM provides only rough suggestions

This is an old way of thinking, and it depends greatly on the granularity and logic of the data used. Today's MMM has the technology to provide more detailed and specific suggestions and enable precise optimization of marketing activities. For example, it is possible to analyze the impact on business results of each consumer segment (new/existing) and sales channel (store/EC). This allows you to quantify how effective each marketing measure is for each, making marketing activities more effective and efficient.

Functionality Misconceptions

8. MMM provides real-time insights

There is a common misconception that MMM provides real-time or near-real-time insights. Although it is possible to speed up analysis by quickly incorporating the latest performance data and by updating analytical models more frequently, the reality is that MMM is more suited to strategic planning than instantaneous decision-making, as it is based on past performance data. MMM analysis is especially valuable for monthly and quarterly planning.

9. MMM is an alternative to other analytical methods

MMM is sometimes perceived as an all-powerful methodology because it can visualize the impact of almost all marketing factors, whether online or offline, as long as there is time-series data, and it can grasp the overall impact of marketing measures. However, in reality, MMM complements other methods and is nothing more than a means of evaluating the effectiveness of marketing from a broader perspective. For example, by combining it with attribution analysis, brand lift research, CMM (Consumer Mix Modeling), etc., you can make more accurate and confident decisions while taking advantage of the strengths of each methodology.

10. AI-based MMM can completely replace human judgment

AI and machine learning are powerful tools in MMM to find hidden patterns from huge amounts of data and improve prediction accuracy. However, AI-based models are often considered "black boxes" because the process and behind-the-scenes of the results cannot be seen. For example, not being able to understand the basis of the results is a big risk when it comes to important strategic decisions such as annual budgets and quarterly budget allocations. In addition, the objective suggestions provided by MMM are based on past data and cannot always keep up with rapid changes in the market and consumers. That is why it is important not only to utilize the results of AI, but also to combine them with marketers' insights and understanding of consumer behavior, correctly interpret the results, and link them to the next action. The fusion of data and human judgment enables more balanced decision-making.

Why do these misconceptions arise?

There are several key factors behind the widespread misunderstanding of MMM, and understanding these factors can provide clues on how to properly understand and use MMM more effectively.

A persistent image of the past

MMM was originally created as a method for large corporations. Globally operating companies have a history of using it to measure the effectiveness of marketing activities across multiple channels, such as television and newspaper advertising. As a result, the image of MMM being a high-cost tool only for corporations with large budgets persists to this day.

Excessive expectations for real-time and short-term responses

Modern marketing is increasingly requiring real-time data analysis. One example is a short-term response such as instantly measuring the effectiveness of an online campaign and reflecting it in the next day's measures. However, MMM is a method for analyzing medium- to long-term trends based on past data, and is not suitable for real-time decision-making. Therefore, excessive expectations of real-timeness can easily lead to the misunderstanding that "MMM is not suitable for modern marketing."

Lack of expertise

MMM requires specialized knowledge because it makes use of advanced statistical methods such as regression analysis and time-series data analysis. For example, it includes many mathematical models and data processing techniques that marketers rarely encounter on a daily basis. For this reason, it is not easy to understand the full picture of MMM, and this complexity is one of the reasons for its misunderstanding. In addition, basic knowledge and experience of statistics are required to correctly interpret the results of MMM and link them to appropriate marketing measures. Without these, the value of MMM cannot be fully realized, which ultimately leads to doubts about its effectiveness.

Eliminate misconceptions and recognize the diversity of MMM

MMM is widely known as a methodology for measuring the effectiveness of marketing activities and optimizing budget allocation, but there is no one-size-fits-all methodology or approach, and MMM can be used in many different ways depending on the specific needs of a company or organization.

For example, some companies develop their own MMM in-house. In this case, the company forms a team of data analysts who are familiar with their own data and MMM to build a customized model optimized for the specific business needs. This approach is labor-intensive, but very flexible and suitable for addressing the unique challenges of companies. On the other hand, MMM provided by advertising agencies and consulting companies is generally a packaged solution based on industry-wide best practices. There is also human support, which allows for quick implementation, making the barrier to implementation low, especially for companies that are new to MMM.

Furthermore, for MMM, MetaRobynor GoogleLightweight MMMUsing these libraries, developers and data scientists at companies, advertising agencies, consulting firms, etc. can improve accuracy at low cost by comparing with in-house developed models.

In addition, XICA providesMMM Solution "MAGELLAN"is a unique MMM with specific technologies and customizability, backed by a track record of supporting over 250 companies. MAGELLAN offers flexible models tailored to the specific needs of companies, and its customizability sets it apart from other MMMs.

In conclusion

MMM is a powerful method for understanding the big picture of marketing activities. By correctly understanding and utilizing MMM, you can build a more effective marketing strategy and achieve sustainable business growth. However, misunderstandings about MMM can be an obstacle to fully realizing its value in the field of marketing. We hope that this article has cleared up such misunderstandings and helped you understand the practical and strategic suggestions that MMM provides.

XICA has over 10 years of experience providing services in MMM and has supported many clients, mainly domestic enterprise companies. If you are interested in learning more about MMM and how it can be applied to your marketing activities, please feel free to contact us.Contact usOur analysts and consultants with extensive expertise in a wide range of industries will help you get the most out of MMM and help many companies and marketers make better decisions.

↓ Click here for a list of carefully selected articles to deepen your understanding of MMM