top of page

Delivered on November 17, 2025

The Relationship Between POIs and Generative AI – And Now to Geofenced POIs! by Digital Advantage Inc.

Speaker: Takahisa Ogawa
CEO, Digital Advantage Inc.

Affiliation and position information is as of the time of distribution

We spoke with Ogawa of Digital Advantage, operator of the chain store data service "Locesma," about utilizing POI data in the age of generative AI. Locesma is a free chain store search service that crawls data from approximately 10,000 chains and 1.14 million stores daily, manually assigning latitude and longitude coordinates to maintain high quality. In recent years, demand has grown in niche categories such as cram schools, family funerals, and secondhand stores. While touching on the threat of generative AI outputting POIs, Ogawa spoke about the importance of data assets held by Japanese companies. As a new initiative, he explained a new concept: Organizing stores as building polygons (geofences) rather than points, and combining this with foot traffic data and official apps to analyze store visits, optimize promotions in real time, and provide highly accurate input data for AI.

====

 

Locesma's High-Precision POI Data and Business Infrastructure

 

Locesma is a free service that allows users to search for chain stores on a map. Its business relies on providing highly accurate store data displayed within the service.

Commitment to Maintaining Data Scale and Quality

 

・Data Scale: As of October 30, 2023, we have collected data on 10,705 chains and 1,138,510 stores.

 

・Freshness and Update Frequency: All data is collected daily via web crawling and thoroughly updated with the latest information.

 

・Manual Quality Control: For new stores added daily (100-500 per day), latitude and longitude are manually set by hand while viewing maps, without relying on geocoding. This maintains high data accuracy and differentiates us from other companies.

 

・Uses and Service Experience: We have provided data to major platforms such as Apple Maps, LINE Yahoo! JAPAN, and KDDI Location Analyzer. Another major source of revenue comes from providing data used by chain store development departments for competitive research and other purposes.

 

・Expanding Demand: In addition to traditional retailers like supermarkets and convenience stores, data demand has increased significantly in recent years in niche categories such as cram schools, funeral homes, and used goods dealers.

Data Strategies and Warnings in the Age of Generative AI

 

Ogawa recognizes the current situation in which generative AI like ChatGPT has begun to output specific POI information as a major threat to data providers.

 

Recognition and Challenges of Generative AI

 

・Concrete Threat: Obtaining POI information has become easier; for example, when you ask an AI to "list the nearest convenience stores," the information is output with latitude and longitude.

・Current Quality: Currently, the POI data output by AI is not of a quality sufficient for business use in terms of "completeness, accuracy, and comprehensiveness," but he is wary of the possibility that data quality may improve in the future.

・Protecting Data Assets: To prevent inadvertent learning by AI crawlers, the company consults with clients regarding the interface and data exposure methods when providing data, taking care to protect its POI data, which are its "property."

Importance of Data Assets for Japanese Companies

 

He emphasized that "data" is the most important factor in AI development, and that the high-quality data accumulated at manufacturing and other on-site locations owned by Japanese companies, which have long been world leaders, will be Japan's greatest strength in the AI ​​era. He stressed the need for company-wide awareness and measures to prevent these valuable data assets from being inadvertently stolen by AI or left unused and buried.

Evolving into "Geofence," a Next-Generation Location Information Analysis Platform
Locesma has embarked on a new initiative to evolve beyond providing only "point" information to "polygonal data of building shapes (geofences)."
 

Purpose and Benefits of Introducing Geofences

 

・High-Precision People Flow Analysis: This will enable accurate determination of whether customers are actually inside a store building, something that was previously impossible with point data.

 

・Achieving Stay Analysis: Combining this with foot traffic data enables deeper analysis, such as customer dwell time and whether or not they visit the store.

 

・High-Quality Input for AI: By inputting clear store visitor data, rather than simply passing by, into the AI, we significantly improve its accuracy.

 

・Development Status: We have completed geofencing for commercial facilities and have begun publishing the system on the Locesmart web version. We are currently working on geofencing for three major convenience stores, with plans to expand to other retail sectors such as supermarkets and drugstores.

Real-Time Utilization Made Possible by Geofencing

 

・Promotion Optimization: We will analyze the effectiveness of nationwide promotions in real time using store visit data, leading to the creation of a system that automatically and instantly changes and optimizes advertising messages and creatives from the next day based on the results.

・Official App Integration: Smartphone apps will use geofencing to accurately track customer in-store time and can automatically switch to in-store mode or award visit points, maximizing customer experience and marketing effectiveness.

Summary: Data Quality and Depth Are Key in the Age of AI

 

Digital Advantage's LocaSma continues to provide irreplaceable, highly accurate POI data, even in the age of generative AI, through thorough manual quality control and daily updates.

Furthermore, the evolution from "points" to "geofences" enables real-time use in people flow data analysis and official app integration, and is a strategy that provides highly accurate "location" information as input for AI utilization. This initiative goes beyond simply providing data; it differentiates itself through the quality and depth of its data, greatly expanding the possibilities of the location business.

bottom of page