XICA MARKETING SCIENCE LAB (hereinafter referred to as MS Lab) is a community that aims to explore "optimal decision-making" in marketing. A free event exclusively for members was held at the Tokyo venue on Thursday, November 2025, 3.

The event was divided into two parts: a "seminar" where cutting-edge information and success stories on data-driven marketing were shared, and a "lab" where practical knowledge was generated through two-way communication between the seminar speakers and participants. This article reports on some of the content of the seminar.

Speakers/panelist:

*Titles are as of the time of the event.

Mr. Ken Nakajima

Executive Officer and Head of Marketing, Kewpie Corporation

Graduated from the Faculty of Agriculture, Tokyo University of Agriculture and Technology in 1993. After joining Kewpie Corporation, he worked in research and development, mainly on prepared foods. Since 2011, he has led the company's overseas expansion strategy from the product development and marketing aspects, supporting the expansion of various bases. After serving as president of the Thai subsidiary from 2018, he will return to Japan in 2021 and work to improve the accuracy of Kewpie's marketing, mainly in the research and development departments. In October 2024, he will become the head of the newly established Marketing Division.

Moderator

Yoshiaki Hirao

President and CEO of XICA Corporation

Graduated from Keio University Faculty of Policy Management. After his father went bankrupt, he strongly wanted to "eliminate the hopeless sadness in the world." Discovering the potential of statistical analysis as a form of management support while at university, he founded XICA Co., Ltd. in February 2012. Based on statistics and economics, he has provided consulting to many major clients on improving marketing accuracy. With this experience as a background, he has served as product owner for the development of various tools at XICA.

table of contents

Kewpie's Executive Officer Talks about the Challenge of Marketing Digital Transformation

Kewpie, which boasts a high market share in mayonnaise and dressings, is taking on the challenge of marketing digital transformation in order to strengthen its core products and expand its value-added products. With favorable results, including a record-high annual profit forecast for this fiscal year, the company is working on organizational collaboration and overall marketing optimization based on data utilization in order to achieve further growth.

For this event, we invited Nakajima Ken, Executive Officer and General Manager of Marketing at Kewpie, to speak about their efforts toward overall optimization of marketing, focusing on the example of their Deep Roasted Sesame Dressing, which has maintained the No. 20 position in the dressing rankings for over 1 years.

The first step in DX: Aiming for data-driven organizational collaboration

Kewpie offers many products that are loved by many consumers, such as mayonnaise and dressings. However, the company had an issue with its marketing in that it was unable to "achieve company-wide optimization as a unified organization." While various factors such as price, shelf space, and advertising affect sales, the impact of each factor could not be measured, making it difficult to reach a common understanding.

Nakajima:"With separate departments such as advertising, sales promotion, product planning and research, each one is working hard on their own assigned tasks. However, there was no basis for aligning perspectives in order to reach a consensus, so each was stuck at individual optimization. To break through this situation, we thought we needed facts such as numbers and data as a common language."

Therefore, Kewpie decided to foster a DX culture within its marketing headquarters in order to promote DX across the entire company. As a first step, they began building a marketing infrastructure. Taking into account the factors mentioned above, they worked on the following to structure and visualize their entire marketing activities:

First, they carried out an analysis aimed at optimizing the quantity of marketing by clarifying "where and how much" to approach. By utilizing MMM (Marketing Mix Modeling), they were able to visualize the contribution of each measure to results from advertising data and performance data, aiming to achieve optimal budget allocation. Next, they carried out an analysis aimed at optimizing the quality by clarifying "who and what" should be delivered. Using CMM (Consumer Mix Modeling), they aimed to elucidate the purchasing mechanism from consumer awareness data and maximize the probability that their brand would be chosen.

The company believed that by implementing this PDCA cycle of quantity and quality, different organizations would be able to align their perspectives based on data, which would lead to overall optimization of marketing.

Nakajima:"At Kewpie, we started by optimizing quantity. Although we had accumulated usable data from the numerous measures we had implemented up until that point, we felt that we were not making full use of this data and were unable to make fact-based decisions. After accumulating knowledge through optimizing quantity, we then decided to move on to optimizing quality. Currently, we are working on the multiplication of quality and quantity."

Analysis examples for total marketing optimization

This event focused on Kewpie's "Deep Roasted Sesame Dressing" from among the many products available, and specific analysis examples were shared.

First, in order to optimize the amount of MMM,MAGELLANMMM is effective in visualizing the contribution of each initiative to results and promoting budget optimization, but Kewpie's initiative has two unique features.

The first is that it takes into account "brand equity (brand accumulation effect)." It quantifies not only "direct effects," "indirect effects" that have a ripple effect on other measures and KPIs, and short- to mid-term "residual effects," but also the "brand accumulation effect" that influences results over the long term. Since Kewpie has been running TV commercials that are similar to brand appeal for a long time, the idea is that by visualizing the brand accumulation effect, it will be possible to correctly grasp the contribution to sales. The other is that it built an analysis model based on the fact that "Deep Roasted Sesame Dressing" is a deep product with a wide range of uses. Because it has a wide range of uses, such as being able to be used in dishes other than salads, it was hypothesized that recipe sites would influence the consideration step.

As a result, the following became clear:

First, we compared the sales amount and ROAS of each campaign, and found that TV commercials (time) had the highest sales amount in both the short and long term. Regarding ROAS, if we focus only on short-term effects, online campaigns (combined) are the best, but when long-term effects are taken into account, TV commercials (spot) were the best. In other words, we found that online campaigns are campaigns that lead to instantaneous results, while TV commercials are campaigns that continue to accumulate results over the long term.

They were also able to obtain suggestions for improvement. For example, TV commercials (time) had the highest sales amount and a good ROAS, but they found that they were far beyond the saturation point. In order to improve efficiency, they could consider a policy of continuous advertising within the optimal investment range.

In addition, when analyzing online initiatives in detail, it was found that recipe sites had the largest long-term sales. As hypothesized, it can be inferred that recipe sites are visited by many people considering dressings.

Next, in the analysis of CMM that leads to quality optimization, we will analyze the mechanism of loyalty and optimize the strategy.COMPASSAt this event, an analysis was presented to explore the drivers that lead to light to middle-heavy players, in other words, loyalty.

The royalization analysis revealed four main findings:

・An effective factor for increasing loyalty is "frequency of use in dishes other than salads."

・Kewpie's image of being "unfailing" and "traditional" is also effective in boosting its "infallible" image.

・Overall, the driving force is the evaluation of the "deep sesame flavor."

・The health-conscious feature of being "low in calories" is also effective

Hirao:"What we want you to pay attention to is the switching rate. The switching rate is the probability that a customer will become loyal as a result of a one-point increase in the survey response score for that item, and it clarifies key drivers. By determining what needs to be promoted from a strategic perspective, it becomes possible to implement measures with a clear focus."

It is believed that it would be effective to aim for loyalty by spiraling up the two axes that the analysis found could be drivers of loyalty: "use in dishes other than salads" = "can be used in a wide range of applications" and "deep-fried sesame flavor" = "like the flavor." This strategy has actually been used in new TV commercials that appeal to "uses other than salads" and in the development of new products in the "deep-fried sesame dressing" series.

Nakajima:"Even before using COMPASS for our analysis, there were some items that we knew we should focus on based on our research and hypotheses. However, because we have been doing this for so many years, it has been difficult to make choices, and we have sometimes invested too much. Now, we have found that there are drivers that are important but do not require as much energy, and we feel that COMPASS is valuable in helping us narrow down where we should invest."

Action optimization using MMM×CMM

Looking back, the results of both MMM and CMM show that "recipe sites" will be a key driver. So Kewpie strengthened its recipe sites. As a result, the contribution of recipe sites to results increased significantly from 2.0% to 4.9%. In addition, since it is assumed that appealing to the use of recipes as a "secret ingredient in cooking" is effective, it seems that they are also considering disseminating related content.

Going forward, the company will continue to optimize quantity and quality, aiming to improve ROAS, strengthen loyalty, and achieve company-wide optimization across the entire organization.

Looking back on the efforts made to date, Nakajima spoke of his feedback and expectations as follows:

Nakajima:"Just because we used MAGELLAN and COMPASS doesn't mean we'll get the answers. But we feel that we've created a common language for finding answers, and that we've laid the starting line. Dressings, among other seasonings, require a wide range of options and originality, so it's essential to regularly release new products and offer a wide variety. That's why it's important not to leave it to chance, but to increase reproducibility and the rate of success. For that reason, I think it's important that we continue to work on our analysis."

To summarise the event, Hirao said, "As Nakajima-san said, marketing is not so simple that answers can be derived from data science alone. However, I believe that by basing it on data, everyone can be on the same page, and overall optimised marketing can be achieved. This will make it possible to devise strategies and tactics with a high success rate that do not rely on KKD (experience, intuition, courage). Through this Kewpie example, I hope that you will see that data-driven optimisation of quantity x quality can make it possible to continue winning with a high degree of certainty."

Recommended articles for those who read this event report

[Event Report]

Optimizing communication by clarifying the structure of consumer awareness and business results

Read the report

[Marketing Agenda Tokyo 2024 Event Report]

Kewpie takes on the challenge of total marketing optimization through digital transformation ── The key to organizational collaboration is a data-driven decision-making platform

Read the report