The Evolution of "Oi Ocha" to Survive the Era of Inflation: Utilizing AI and Data Science to Assetize "Decision Making" [Marketing Agenda 2026 Event Report]
Ito En is facing headwinds such as soaring raw material costs and the "Japanese tea crisis" stemming from the global competition for high-quality tea leaves. To overcome this situation, their best-selling brand, "Oi Ocha," is shifting from a defensive to an offensive, data-driven strategy.
On June 3, 2026, at "Marketing Agenda 2026," Japan's premier marketing conference held in Okinawa, Mitsumasa Shida, Executive Officer and Head of Marketing at Ito En Co., Ltd., and Yoshiaki Hirao, President and CEO of XICA, took the stage. In line with the event's theme, "The Evolution of Marketers," they revealed the inner workings of the "evolution of decision-making" required in the age of AI. They unveiled the entire practical process of how they are using AI and data science to transform decision-making and lead brands to further evolution.
This report will introduce the contents of the sessions held on the day.
*Titles are as of the time of the event.


table of contents
- What kind of "evolution of marketers" is required in the age of AI?
- The "Japanese Tea Crisis" We Face and Ito En's Commitment
- CORE 01: Semanticizing Data | Unraveling the Royalization Driver
- CORE 02: Advanced Decision Making | AI-Powered Scenario Analysis and Countermeasures
- CORE 03: Assetizing Organizational Knowledge | Future Evolution through Decision-Making Management
- Ultimately, it is "people," not AI, who will take the helm.
What kind of "evolution of marketers" is required in the age of AI?
At the beginning of the session, Hirao said"In an era where AI utilization becomes commonplace, what will determine success or failure is the 'accuracy and speed of decision-making' of the people and organizations using AI—in other words, the 'evolution of decision-making.'"He stated that while many organizations are making progress in using data and AI, the lack of fundamental changes in decision-making is highlighting the following challenges.
Challenges in decision-making
- The basis for the judgment is vague.Even if we gain insights from AI, the interpretation is left to each individual, and ultimately we end up relying on our own judgment based on "KKD (experience, intuition, and courage)."
- Our ways of thinking and perspectives are not aligned.The issue remains unclear, actions are taken prematurely, and ultimately, the situation becomes one where one party has to wait for a final decision.
- Organizational knowledge is not accumulated.The process of "why that decision was made" is not recorded, and know-how and experience remain confined to individuals.
Thus, unless the decision-making process itself evolves, no matter how much we invest in data and AI, the results will plateau.
To address this challenge, XICA is currently implementing a new system that integrates analysis, decision-making, execution, and learning into a single, seamless process."XICA*AIP (AI Platform)," a proprietary AI platform that systematizes the decision-making process.XICA*AIP works in conjunction with "three value propositions" to transform data into a business lever and elevate organizations into a "continuously winning structure." It transforms mere data into evidence-based, core information."CORE 01: Semanticization of Data (Decision Insight)"To understand the organization's unique context and align dispersed thinking and arguments."CORE 02: Decision Refinement"And create a loop where the organization continuously learns by accumulating decisions and their results."CORE 03: Assetizing Organizational Knowledge (Decision Intelligence)"Those are the three.
In the next chapter, we will introduce the details of these core elements, which go beyond simply improving efficiency in one area and maximize the "precision and speed of decision-making" across management and business as a whole, along with our collaboration with Ito En.

The "Japanese Tea Crisis" We Face and Ito En's Commitment
Mr. Shida of Ito En expressed his concern about the current market environment, which is a complete departure from the deflationary market of the past, saying the following:
"As has been widely reported in the media recently, the price of tea leaves has skyrocketed several times over in the last two years. Furthermore, against the backdrop of the global matcha and Japanese tea boom, high-quality domestic tea leaves are being bought up by foreign capital, and in their place, cheaper foreign substitutes are flooding the market and occupying the domestic market. This is a serious situation that could be described as a 'loss of culture,' a 'Japanese tea crisis' (Mr. Shida)."
Amid these headwinds, some competitors are beginning to use imported tea leaves (extract powder) to reduce costs,Ito En is committed to its brand pride of using "100% domestically produced tea leaves," promoting overwhelming differentiation through quality.They adopted this policy. However, in order to stick to that principle, they needed to resolve several organizational issues.
Simply passing on the increased costs due to quality assurance to consumers through price increases will not be accepted. That's why,Eliminating the black box of marketing and enhancing the organization's "tangible" feel.about,Scientifically unraveling the essential reasons (drivers other than price) that consumers "choose"To do, andDevelop strategies that can flexibly respond to uncertain market changes.It had become an urgent matter to do so.
CORE 01: Semanticizing Data | Unraveling the Royalization Driver
The first thing we did was,"CMM (Consumer Mix Modeling)Unraveling the Key Drivers in Customer Loyalty for "Oi Ocha" Green Tea PET Bottles.
The data used in CMM comes from consumer surveys. It meticulously gathers information on the attributes of both moderate and heavy users of "Oi Ocha" green tea PET bottles, their social recommendations, and their perceptions of functional and emotional value. Typical surveys tend to lead to the simplistic conclusion that items with high average scores (i.e., items with good current scores) should be prioritized and strengthened. However, CMM applies a unique and advanced approach combining statistical models and machine learning, aiming to go beyond simply identifying "which elements are high" and instead focus on...This research aims to uncover the true key drivers that transform customer behavior and their probability (switch rate), specifically identifying which factors, when improved, lead to a shift from a mid-level user to a heavy user.To do.
The results of this thorough investigation into the underlying mechanisms overturned conventional wisdom and established assumptions.
"Taste" was not the key driver of the royalization.
- The truth revealedAs a factor in turning moderate users into heavy users,One of the hypotheses raised within the company was that "delicious taste/compatibility with food" is the lifeblood of a brand.However, the analysis revealed that "taste" was not the direct driver of the royalization.What truly made customers loyal was "purchase of other products (a wide variety of products)."The result was that "purchase of other products" had a low average score in the survey, suggesting that it was a key driver that would likely have been overlooked using conventional methods.
- The Chasm of HeavinessThe study visualized a correlation where the proportion of heavy users begins to increase when the number of purchased items exceeds "4," and surpasses 7% when it reaches "5."
- Effective Shopping StepsIt was found that the key to becoming a loyal customer lies in how to bridge the gap from the first product, "Oi Ocha" green tea PET bottle, to other products in the series such as "Koicha," "Hojicha," and "Genmaicha," and then to the fifth product, "Trial" (including loose leaf and matcha products, and PURE GREEN, which will be released in 2025).

Furthermore, it is the first thing customers think of when they think of that brand.The key factor in boosting "first-recall" is "understanding the characteristic of being the No. 1 seller."It was also identified that this is highly correlated withThe brand consistently promotes images such as "a pioneer of green tea" and "drinkable all over the world."This led to the hypothesis (structure) that this approach is effective in building a convincing royalization structure.

Shida reflects on the impact of this analysis as follows:
"This analysis brilliantly overturned our preconceived notions as a manufacturer. We tend to think, 'Our products are good,' or 'They sell because they're delicious.' However, from the customer's perspective, deliciousness is already a given. Increasing the number of products they buy and try is what ultimately leads to customer loyalty."What we should really be doing is not simply selling green tea, but enriching our customers' lives through the diverse variety of teas we offer.This initiative has made me realize once again that this is indeed the case (Shida)."

CORE 02: Advanced Decision Making|AI-powered scenario analysis and countermeasures
The focus then shifted to how to respond to rapidly changing market conditions. Even in a market with increasing uncertainty,To enable timely decision-making and action in response to change, AI is used to comprehensively visualize possible scenarios.I tried it.
The AI generated a staggering 191 different strategic simulations.This is the case. Based on the structured mechanism in CORE 01, it covers a range of scenarios, from positive scenarios where added value is accepted to negative scenarios such as customer defection due to a return to low prices or a shift to private brands, and category contraction scenarios due to the spread of reusable bottles. By anticipating risks, opportunities, and recommended actions in advance, a system has been built that can maximize the speed of decision-making even in uncertain situations.

In this simulation, the most common problem that emerged was:The limitations of the functional value proposition we have emphasized so far, such as "100% domestically produced tea leaves," and the conservative context of "it's good because it's domestically produced."It is.
Therefore, Ito En,We made the decision to strategically elevate our existing appeals and build a new brand story.Specifically, it goes beyond simply appealing to quality, aiming for a socially responsible narrative that seeks to revitalize tea-producing regions in a sustainable way, working with producers as a "pioneer of green tea." Furthermore, it combines the strength of being consumed all over the world with a unique "Champagne concept" that elevates Japanese tea to an internationally recognized refinement.Building "regional branding" to drive next-generation global branding.They changed course.
This new strategy, forged through thorough collaboration with data science and AI, led to the press conference in May 2026. Instead of simply proclaiming "100% domestically produced tea leaves," the company announced strengthening its "New Tea Producing Area Development Project," which involves working with producers to create actual tea-producing regions, and the "National GI (Geographical Indication)* Japanese Tea Concept," aimed at enhancing the international value of Japanese tea.
Shida says that while working on this project with XICA, they arrived at a decisive conclusion.
"After examining various scenarios, we came to the conclusion that a strategy that simply ensures the survival of our own brand will ultimately not lead to continued survival."What we should be aiming for is to "grow the entire category of tea."Using inexpensive raw materials from overseas would eliminate the high cost, but that's not the solution.We declared our commitment to supporting domestically produced tea, developing the tea industry together, and enhancing the value of the category."That's why. All of this was possible because we thoroughly discussed and brainstormed with XICA, and although we had doubts along the way, we all finally came to terms with it. I am confident that it was by utilizing data science and AI that we were able to take a step that went beyond the conventional framework of a manufacturer (Shida)."
The announcement generated a huge response, with over 200 media outlets flocking to the event in just two days. It was a resounding success as an extremely unusual and socially impactful initiative that defied conventional wisdom.
*A system to protect the names of agricultural, forestry, and fishery products and food products as intellectual property, which have a tradition nurtured in the region and whose high quality is linked to the place of production.

CORE 03: Capitalizing organizational knowledge|Future evolution through decision-making management
Regarding the third core principle, "assetizing organizational knowledge," Hirao presented an implementation demonstration.
In actual operation, for example, when asked "Why are sales of this product declining?", the AI correctly understands the underlying context and immediately searches for related factors. It then identifies the root cause, such as "a decline in top-of-mind recall," and visualizes which measures are contributing to that decline in a chain reaction. This system allows for a smooth transition to the ideation phase by repeatedly engaging in dialogue and brainstorming with the AI, deepening the understanding of the problem, and prioritizing measures.
As the analysis progresses and specific issues such as "the asset value of creative work is declining" are identified, the AI will suggest optimal creative improvement plans tailored to the target audience. The team can also submit sophisticated, data-driven ideas—not just random thoughts—directly to managers through the platform (requesting approval). All of these decision-making logs, as well as the results of currently implemented initiatives such as sales contributions, are stored as data within the organization on the platform.
Hirao said, "ThisThe accumulated organizational knowledge (CORE 03) is then fed back into the next analysis (CORE 01) and the refinement of discussions (CORE 02). Continuously turning this spiral and raising it is essential for "continuing to win."He emphasized its significance, saying, "It is so."

Furthermore, among the three coresThe most effective ways to differentiate yourself from competitors are "CORE 01," which transforms data into unique meaning, and "CORE 03," which stores your own decision-making processes and their results.Hirao mentioned this, and argued that CORE 03 offers significant first-mover advantages, meaning that the sooner you start using it, the greater your competitive edge will be.

Ultimately, it is "people," not AI, who will take the helm.
To conclude the session, Hirao forcefully declared XICA's clear stance on data science and AI.
"There is absolutely no way AI will lead to victory."The future we truly await is not a victory achieved by AI, but a victory achieved by humans using AI to its fullest potential.That's it. Instead of entrusting decision-making to AI, we will make decisions based on our own will while making full use of AI.To implement the "evolution of decision-making" within the organization, we will connect the analytical power we have honed to date to the next level and define ourselves asAI Decision EngineRedefined as 'We have taken the first step in this direction. We will provide a decision-making platform that will allow marketers and management to confidently take their next move (Hirao)."
Mr. Shida also expressed his expectations for the worldview that data science and AI will bring.
"What was so great about working with XICA-san was that in every project, we were able to accumulate experience and organize various stories and brainstorm ideas. Above all,In the decision-making process—who decides what—it's important that all members fully understand and agree with the decision before proceeding.We were able to achieve this. Furthermore, because we have 191 different scenarios and countermeasures at our disposal, even if the market environment differs from our expectations, we can immediately think of and implement the next steps without fear. This approach will further accelerate the speed at which we improve our brands and businesses. We hope to continue to invigorate Japan's tea industry and market with this wonderful system in place (Mr. Shida)."
Rather than a future where AI automatically provides answers, we aim to create a future where humans confidently take the next step through thorough dialogue with AI. What data science and AI truly update is not operational efficiency, but the very "decision-making" of organizations that move forward without fear of uncertainty.

Related Articles
-
About XICADataScience for Growth: Marugame Seimen's Method for Maximizing "Winning Probability" Through Data Science – The Key to Decision-Making That Overcomes Uncertainty [MarkeZine Day 2026 Spring Event Report]
-
About XICAST Corporation's Challenge: Redefining Mature Brands Through "Data Science x AI x Human Resources" [Brand Summit Spring 2026 Event Report]
-
About XICAMaximizing the odds of success through “sensitivity x data science”──A look behind the scenes at Marugame Seimen’s unstoppable progress [Marketing Agenda 2025 Event Report]


































