Case Study: Inquiry Chat AI "miibo" Addresses "Small Questions" on Money Forward's LP, Detailed Interview Released on miibo Official Website
miibo Co., Ltd. has released a case study detailing the implementation of its AI chatbot "miibo" on Money Forward's corporate cloud service landing page (LP). Leveraging advancements in generative AI, miibo successfully resolves users' "small questions" on the web, achieving approximately an 80% inquiry resolution rate and demonstrating the effectiveness of web-based customer service AI chat.
📋 Article Processing Timeline
- 📰 Published: April 24, 2026 at 20:15
- 🔍 Collected: April 24, 2026 at 12:01
- 🤖 AI Analyzed: April 25, 2026 at 01:37 (13h 35m after Collected)
miibo Co., Ltd. (Headquarters: Minato-ku, Tokyo, Representative Director: Masashi Kunugi) has released an interview on its official website with Hiromi Watari of Money Forward's Cross-Segment Marketing Division, Organic Growth Department, regarding the case study of AI chatbot utilization on the corporate service website for "Money Forward Cloud," a cloud service covering all back-office operations such as accounting, invoicing, expenses, and payroll, provided by Money Forward Co., Ltd.
**Key Highlights**
* Realized a "web-based customer service type" AI chat that resolves "small questions" directly on the LP.
* After two previous setbacks, a re-challenge and full-scale adoption with miibo was pursued, driven by improvements in generative AI accuracy.
* Ensured high response accuracy by separating pricing information management (via prompt) from other knowledge management.
* Established a continuous knowledge improvement cycle by integrating conversation logs with Slack and summarizing them with AI.
**Main Content of the Interview Article**
**Background for chatbot implementation**
* For Money Forward Cloud's corporate LPs, which offer multiple back-office SaaS, a challenge was responding to "small questions" regarding complex pricing structures and applicability to specific company conditions.
* There was no free-word search on the site, necessitating a means for users to resolve questions on the spot.
* After two failures (pre-generative AI era and repurposing product-specific chat), a re-challenge was made due to improved generative AI accuracy: "We thought it might work now that generative AI has improved, so we started considering tool adoption again," said Watari.
**Reasons for choosing miibo**
* While searching for a tool better suited for "web-based customer service" rather than "inquiry reception," consideration began following an internal use case for another service at Money Forward.
* Deciding factors were the flexibility to switch LLMs freely, low implementation cost, and comprehensive Japanese support: "I was a bit uneasy about committing to a single model. The ability to choose models as needed was reassuring," said Watari.
**Ingenuity in setup and operation**
* Pricing information, which requires accuracy, is managed on the prompt side, while other information is obtained from the knowledge base.
* LP expressions with variations (e.g., "Use for one month free" and "Start free trial here") were organized for AI to handle without discrepancies.
* From decision to release took 1-2 months; setup was handled by the person in charge, and LP implementation by engineers.
**Concrete results from implementation**
* Achieved approximately 80% question resolution rate by AI on the pricing page.
* During peak seasons (e.g., tax filing season), thousands of unique users per month utilized the chat.
* "Small questions" that previously didn't surface through traditional inquiry forms became visible, providing hints for LP improvement: "Voices asking for minor clarifications, which wouldn't have been asked through inquiries, become visible when interacting with AI," said Watari.
**Future Outlook**
* Advance data preparation and strengthen real-time actions for unanswered questions (e.g., presenting online reservation guides for highly prospective questions).
* Explore ways for AI to handle supplementary information not fully covered on the LP and individual confirmations, considering changes in user information gathering behavior (e.g., increased cases of users narrowing down options with AI before visiting the LP).
**Significance of the Interview**
* A concrete practical example of AI chat utilization for "pre-conversion touchpoints" in B2B SaaS.
* An example of AI design balancing accuracy and operational efficiency through the division of roles between prompts and knowledge.
* Presentation of a new approach combining marketing and AI, namely an LP improvement cycle using conversation logs.
* Redefinition of AI chat usage as "web-based customer service" rather than "inquiry handling."
**Read the full interview here**
Detailed interviews can be found in the article below.
miibo Official Website: https://miibo.ai/interview/moneyforward
**About miibo Co., Ltd.**
miibo Co., Ltd., with the mission "To enrich people's lives with AI technology," develops and provides the conversational AI building platform "miibo." As shown in this case, it is committed to solving societal challenges through AI technology.
https://miibo.co.jp
**About the conversational AI building platform "miibo"**
"miibo," the conversational AI building platform developed by miibo Co., Ltd., is a service that allows users to easily create practical conversational AI without code.
Official Website: https://miibo.ai
**Easy AI Application Creation for Everyone**
No difficult skills or languages required! It's a conversational AI building platform that enables you to quickly create AI-powered applications utilizing your existing databases and large language models (LLMs).
**Build and Test with Super Agile Development**
Rapid practical creation and testing.
**Key Highlights**
* Realized a "web-based customer service type" AI chat that resolves "small questions" directly on the LP.
* After two previous setbacks, a re-challenge and full-scale adoption with miibo was pursued, driven by improvements in generative AI accuracy.
* Ensured high response accuracy by separating pricing information management (via prompt) from other knowledge management.
* Established a continuous knowledge improvement cycle by integrating conversation logs with Slack and summarizing them with AI.
**Main Content of the Interview Article**
**Background for chatbot implementation**
* For Money Forward Cloud's corporate LPs, which offer multiple back-office SaaS, a challenge was responding to "small questions" regarding complex pricing structures and applicability to specific company conditions.
* There was no free-word search on the site, necessitating a means for users to resolve questions on the spot.
* After two failures (pre-generative AI era and repurposing product-specific chat), a re-challenge was made due to improved generative AI accuracy: "We thought it might work now that generative AI has improved, so we started considering tool adoption again," said Watari.
**Reasons for choosing miibo**
* While searching for a tool better suited for "web-based customer service" rather than "inquiry reception," consideration began following an internal use case for another service at Money Forward.
* Deciding factors were the flexibility to switch LLMs freely, low implementation cost, and comprehensive Japanese support: "I was a bit uneasy about committing to a single model. The ability to choose models as needed was reassuring," said Watari.
**Ingenuity in setup and operation**
* Pricing information, which requires accuracy, is managed on the prompt side, while other information is obtained from the knowledge base.
* LP expressions with variations (e.g., "Use for one month free" and "Start free trial here") were organized for AI to handle without discrepancies.
* From decision to release took 1-2 months; setup was handled by the person in charge, and LP implementation by engineers.
**Concrete results from implementation**
* Achieved approximately 80% question resolution rate by AI on the pricing page.
* During peak seasons (e.g., tax filing season), thousands of unique users per month utilized the chat.
* "Small questions" that previously didn't surface through traditional inquiry forms became visible, providing hints for LP improvement: "Voices asking for minor clarifications, which wouldn't have been asked through inquiries, become visible when interacting with AI," said Watari.
**Future Outlook**
* Advance data preparation and strengthen real-time actions for unanswered questions (e.g., presenting online reservation guides for highly prospective questions).
* Explore ways for AI to handle supplementary information not fully covered on the LP and individual confirmations, considering changes in user information gathering behavior (e.g., increased cases of users narrowing down options with AI before visiting the LP).
**Significance of the Interview**
* A concrete practical example of AI chat utilization for "pre-conversion touchpoints" in B2B SaaS.
* An example of AI design balancing accuracy and operational efficiency through the division of roles between prompts and knowledge.
* Presentation of a new approach combining marketing and AI, namely an LP improvement cycle using conversation logs.
* Redefinition of AI chat usage as "web-based customer service" rather than "inquiry handling."
**Read the full interview here**
Detailed interviews can be found in the article below.
miibo Official Website: https://miibo.ai/interview/moneyforward
**About miibo Co., Ltd.**
miibo Co., Ltd., with the mission "To enrich people's lives with AI technology," develops and provides the conversational AI building platform "miibo." As shown in this case, it is committed to solving societal challenges through AI technology.
https://miibo.co.jp
**About the conversational AI building platform "miibo"**
"miibo," the conversational AI building platform developed by miibo Co., Ltd., is a service that allows users to easily create practical conversational AI without code.
Official Website: https://miibo.ai
**Easy AI Application Creation for Everyone**
No difficult skills or languages required! It's a conversational AI building platform that enables you to quickly create AI-powered applications utilizing your existing databases and large language models (LLMs).
**Build and Test with Super Agile Development**
Rapid practical creation and testing.