Clarifying the relationship between consumer awareness and business results, and optimizing communication

KDDI and SAICA began their data-driven marketing initiative in 2019. KDDI, which offers a wide range of services in the telecommunications industry, is using SAICA's MMM (Marketing Mix Modeling) to optimize communications to maximize results in the complex customer acquisition process.

Five years after its implementation, KDDI's use of MMM has progressed significantly, with the company achieving frequent PDCA cycles through monthly analysis*5. UQ mobile, which began analysis in 1, achieved notable results, such as reducing communication costs per subscriber by approximately 2021% compared to before optimization, by optimizing its monthly marketing investment*2.

In this case study, we will introduce the data-driven marketing optimization methods that KDDI is working on with XICA, focusing on a project at UQ mobile.

*1 As of August 2024
*2 FY2023 results

The key to success in a crowded market is to be "data-driven" and "speedy"

The market for communication services such as mobile phones and smartphones has many competitors, and because it is difficult to differentiate in terms of functionality, it is prone to price wars. Therefore, differentiation through communication is important, and for this reason, optimization of creativity and media is required.

Another notable change in recent years is that customers are becoming less conscious of their mobile carriers. More and more people are not interested in carriers and don't care which one they use. Because it is a stock-based business model, KDDI's own survey shows that customers who are in the purchasing phase only make up about 2% of the market.

KDDI is targeting consumers who make up about 8% of the market but have not yet entered the purchasing behavior stage (hereafter referred to as the "everyday stage"), with the aim of helping them form an image of KDDI's telecommunications brands (au, UQ mobile, povo) from the everyday stage and having KDDI be the first name in their mind when considering switching carriers.Implementing measures tailored to the targetWe are trying to

In order to maximize the results of these marketing activities, KDDI places importance on:Data-driven decision making and speed of analysis cycles.

The process by which customers sign up or sign up for KDDI's services, including UQ mobile, is extremely complex and characterized by the intertwining of numerous factors. In order to correctly grasp the influence of these factors and make more appropriate decisions, KDDI visualizes these influences based on large amounts of data. This is the main purpose of using MMM.

One of the features of KDDI's use of MMM is that the analysis is done on a monthly basis. KDDI mainly uses MMM to optimize online media creative and budgets, and performs analysis in line with the monthly operational cycle.

Communications services market environment

Enabling data-driven marketing decisions

As mentioned above, KDDI has continued to optimize its marketing investments on a monthly basis, and as a result has achieved a roughly 2% reduction in communication costs per subscriber compared to before optimization. From here, we will introduce how KDDI makes data-based decisions.

KDDI has four goals in utilizing MMM:

• Accurately measure and visualize the effectiveness of marketing initiatives
• Clarifying the relationship between policies, consumer awareness, and results
• Identify the best approach for each target
• Achieve efficient allocation of marketing budget

The model built to achieve this goal is a major feature of KDDI's MMM.

Features of KDDI's analysis model

• Incorporating consumer awareness indicators into the model to clarify the structure of "measures → awareness → results"
• Set target awareness indicators according to user flow
• Divide results into multiple categories and clarify effective measures for each target

KDDI conducts consumer awareness surveys on a weekly basis, and is able to incorporate the survey results into MMM variables. Attitude indicators that are expected to have a large impact on results are incorporated into the model in line with user paths, making it possible to analyze which awareness indicators should be increased to maximize results, and which measures should be strengthened to increase those awareness indicators. This makes it possible to optimize not only the amount of investment and budget allocation in measures, but also the quality of what awareness indicators should be improved.

UQ Mobile's sales channels include not only brick-and-mortar channels such as stores, but also online channels. By dividing the results, which are the objective variables of MMM, into multiple patterns, we can see which measures are effective in improving each result.

KDDI incorporates consumer awareness indicators into its MMM analysis model

Realization of monthly operation

Another feature of KDDI's use of MMM is the speed of the PDCA cycle, which involves monthly analysis and optimization. The reason why a monthly analysis cycle is necessary is that KDDI's marketing investment optimization is centered on online media, and the PDCA cycle is realized by incorporating new trial measures.

Of course, we optimize our marketing efforts to take into account the effects of offline media, particularly TV commercials, but we also use MMM to improve the efficiency of our marketing, with a focus on online media, which requires quicker evaluation.

Optimization cycle for better results

Monthly MMM analysis cycle performed by KDDI

In running MMM monthly, not only the speed of analysis but also the speed of data collection is important. At KDDI, we collect and integrate marketing data related to our company and competitors from both inside and outside the company, and we create a visualization environment that includes data for analysis and data not used for analysis, ensuring the speed of data utilization. The following two points are important at this time.

• Collect data in different formats and at different granularities and integrate it into data for analysis.
• Build a one-stop tracking environment that allows analysts to grasp abnormal values ​​and trend changes

As shown in the diagram below, a distinctive feature of this system is that it has created an environment in which the results of MMM analysis can be constantly visualized. The creation of such a data infrastructure is the key to data-driven marketing at KDDI.

How to build the data infrastructure that will be the core of KDDI's data-driven marketing

To move the organization

To implement strategies and tactics thoroughly, the involvement of the entire organization is necessary, but generally, the larger the organization, the more difficult this becomes. The Communication Design Department, which Mr. Baba heads, organizes advertising operation projects and works to optimize advertising operations overall while understanding each other's organizations across the group. This not only enables faster operations, but also makes it possible to optimize the entire brand communication in a seamless manner.

In addition, for internal collaboration across organizations, the results of MMM analysis are extremely useful in that they allow the evaluation of each measure to be discussed using common indicators. By evaluating measures side by side using common indicators, it is possible to achieve overall optimization by integrating measures, rather than partial optimization for each measure.

A project to optimize advertising operations

Aiming for further optimization of marketing

As mentioned above, KDDI is using MMM to optimize data-driven marketing investments, and is currently also working on further optimizing creative using new generative AI*.

*This is not the content of the project with XICA.

KDDI is taking on the challenge of optimizing both quality (who and what) and quantity (where and how much), both of which are important in marketing, in a data-driven manner. XICA will continue to accompany KDDI as they continue to evolve in the future.

"XICA's MMM is extremely valuable in that it can perform complex modeling and visualization from large amounts of data. In addition, with the support of consultants, we can achieve a sense of speed from analysis to application of measures, and it has become an indispensable partner for us today."
Mr. Baba, KDDI Corporation