Who are Toyota fans? We used "MINUKERU" to uncover insights from X posts!

In June 2024, Dentsu Inc., in collaboration with Professor Sotaro Katsumata of Osaka University's Graduate School of Economics, launched "MINUKERU" (*), a data solution that extracts "context" from text data that is invisible through conventional analysis ( see overview here ). "MINUKERU" can swiftly extract multiple contexts from text sources like surveys and user reviews while protecting personal information.

This time, we invited Mr. Akihiro Takehara from TOYOTA CONIQ Pro, Inc. Brand Marketing Division, who is actually using "MINUKERU" to analyze X platform post data. We spoke with him about specific applications and the potential of "MINUKERU." The interviewer is Mr. Kazuki Takahashi from Dentsu Inc., one of the "MINUKERU" development team members.

Dentsu Inc. analyzes the data and delivers reports on the results to clients. Based on these analysis results, Dentsu Inc. can also develop action plans.
 

 

Analyzing X Posts with "MINUKERU" to Respond to Changing Automotive Demand

Takahashi: TOYOTA CONIQ Pro, Inc. is a marketing company supporting the transformation of the Toyota Group. Last year, I introduced them to the "MINUKERU" system we were developing. From there, we worked together on the groundwork and conducted an analysis of approximately 15 months' worth of X posts. First, could you tell us why you decided to try using "MINUKERU," the challenges you faced, and the background?

Takehara: I believe this is an issue for the entire automotive industry, not just our company. In recent years, owning a car is no longer a given. More people are choosing not to own cars, while others increasingly use cars through subscriptions or car-sharing services. The very nature of car demand is changing.

Against this backdrop, we believe our marketing activities must focus not just on boosting car demand, but on realizing a "better mobility society" that accommodates diverse ways of owning and using cars.

As part of this, we felt it was crucial to understand the voices of those who are interested in Toyota and actively share information about us. We wanted to value the voices and communications of customers who are already Toyota fans or users who hold a favorable view of us, and leverage that for our marketing and communications.

Takahashi: So that's why you decided to try using "MINUKERU"?

Takehara: Yes. While we had been analyzing X and Instagram posts before, this time we wanted more granular, high-resolution analysis. That's when we were introduced to "MINUKERU," which combines text and numerical data for highly accurate analysis, leading us to decide to try it.

Rather than just vaguely looking at post content or trends, we heard that it could identify users highly interested in Toyota, reveal their posting tendencies based on personal interests, and show their follower counts and influence. We thought, "This is interesting," and decided to implement it.

Analyzing posts × profile information × number of follows/followers reveals user profiles

Takahashi: Specifically, what kind of analysis did you conduct? Could you tell us about the X post analysis you performed using "MINUKERU"?

Takehara: First, we extracted account data for users posting content incorporating Toyota-related keywords like "Toyota," "Alphard," "GR86," and "Toyota Times." At this stage, we had around 500,000 accounts. From there, we filtered users who had posted Toyota-related content at least 20 times over the past 15 months, ultimately narrowing it down to 6,000 accounts.

For these 6,000 accounts, we statistically analyzed "what kind of people with what interests and preferences exist, and what kind of posts they make" using their X profile information and post data. We discovered that even among those posting primarily about cars, there were distinct groups: "Fans who enjoy looking at cars," "Fans who enjoy driving cars," "Fans who collect related merchandise," and "Fans who share car information." For example, "Looking fans" often enjoy motorsports, while "Driving fans" frequently ride motorcycles as well as driving cars.

Furthermore, we observed that "collector fans" frequently collect miniature cars, while "information-sharing fans" tend to actively share new car information and enjoy the reactions it generates.

Takehara: What I found particularly interesting was that each group showed distinct preferences and posting characteristics beyond just cars. For instance, "viewing fans" often showed high interest in cycling competitions alongside motorsports, tending to enjoy both sports while posting live updates. "Driving fans" frequently had hobbies like outdoor activities and camping, while "collecting fans" often combined their passion with niche interests like anime, games, railways, or cameras.

While the frequently occurring keywords and post characteristics differed across groups, we also found that similar vehicle keywords like "GR," "86," and "Supra" appeared frequently across all groups.

Takehara: By carefully examining X posts at this granular level, we gained insight into the profile of Toyota fans. It also became clear what kind of information resonates with each group and what truly benefits the users.

Takahashi: How about follower counts? How did you analyze them, and what did you learn?

Takehara: News-focused accounts run by corporations still seem to have the most followers. Next are "information-sharing fan" accounts run by individuals, which also have high follower counts. This reaffirmed the importance of having these individuals help spread the concept of "life with cars."

Takahashi: It's not just simple observations like "There are a lot of 'viewing fans (accounts)'" or "There are quite a few 'driving fans (accounts)' too." It's more like: "While the number of accounts itself isn't that high, the ones with large follower counts and strong dissemination power are the 'information-sharing news fans (accounts).'" Or: "The ones with high follow counts and follower counts, closely connected with fans sharing similar preferences, are the 'information-sharing individual fans.'" It means we can understand their reach and influence, right?

Takehara: Exactly. This analysis revealed what groups of people with what hobbies and preferences are posting in what contexts, and it even visualized their influence and reach. We achieved a level of detailed post analysis that wasn't possible with the superficial text data analysis we did in the past, and I'm very satisfied.

We want to share information that aligns with users' interests and concerns, not just focusing on cars.

Takahashi: I think X's profiles show personality in a good way. Combining regular posts with this profile information makes the personas of target groups clearer – that's MINUKERU's strength, right?

Takehara: I agree. Because it's text-based and a highly anonymous SNS, it reveals actual personas rather than just formal profiles like "20s male" or "30s female." Analyzing it reaffirmed how interesting X profile information is as an analysis subject.

Takahashi: Thank you. Hearing all this, I'm delighted to learn that MINUKERU contributed analysis results that helped TOYOTA CONIQ Pro, Inc. develop deeper user communication strategies.

This is my personal view, but I think Toyota is fundamentally seen as a good company by most people. Starting from that position, further increasing engagement is quite a challenging task. I felt that traditional methods would struggle to achieve goals like "promoting a deeper level of understanding" or "building stronger networks and communication."

That's why we introduced "MINUKERU" as a tool to spotlight active fans and analyze them with high resolution. I feel it was a perfect fit. I'd be delighted if it continues contributing to creating better user experiences, higher engagement, and 360-degree experiences.

Takahashi: Finally, could you share your plans and outlook for future utilization?

Takehara: Currently, we're exploring what initiatives to implement based on the analysis results. Using these findings as a foundation, we'll determine what information to deliver to which audiences and aim to apply this to actual PR and communication efforts.

Furthermore, as analysis results derived from valuable, raw user feedback, we want to apply them not only to SNS but also to advertising, in-store information provision, customer service, events, campaigns, and other initiatives.

This analysis used approximately 15 months of data. By conducting regular analyses over similarly defined periods, we believe we can also track the growth or decline of each group's accounts and the evolution of trends.

We're now in an era where simply providing car information alone struggles to reach users. The key challenge is how well we can combine this with users' areas of interest, moving beyond just the car itself to deliver information that truly resonates with customers. Moving forward, we aim to leverage "MINUKERU" to provide a user experience that enriches life with cars, ensuring we can effectively respond to the evolving nature of mobility.