APC 技術ブログ


株式会社 エーピーコミュニケーションズの技術ブログです。

How McDonalds Uses ML to Optimize Restaurant Site Selection


As McDonald's actively pushes to expand its global presence, the company utilizes advanced machine learning technologies to smartly determine potential new store locations. This segment delves into how mathematical models and sophisticated data processing refine these critical decisions, ultimately enhancing profitability across the entire network.

Using Optimization Models

McDonald’s emphasizes not just the profitability of individual locations, but also the collective benefit to its entire network. By leveraging professionally constructed "optimization models," these models go beyond mere speculation. They incorporate sales forecasts and consider various factors such as construction and ongoing operational costs. These models are specifically designed to assess potential financial contributions to McDonald's extensive network.

Feature Engineering Process

The feature engineering stage plays a crucial role in constructing these optimization models. This process involves carefully selecting and preparing predictors from a vast range of data sources. These factors are then used as inputs for the optimization models, enhancing the accuracy of site selection predictions and the overall decision-making process.

This session highlighted how large corporations like McDonald’s utilize cutting-edge machine learning tools and data analytics technologies for strategic decision-making. Advances and refinements in these technologies remain a central focus as the company moves forward.

How McDonald's Leverages ML to Optimize Restaurant Site Selection

McDonald’s adopts a sophisticated process that involves acquiring specific geographic coordinates (longitude and latitude) and feeding them into ML models to optimize site selection. This advanced location-decision capability surpasses what many companies can achieve.

A notable highlight from the session was the discussion on scaling by the speakers. They mentioned, "After deploying end-to-end solutions, support needs to be expanded across multiple markets." They emphasized the importance of appropriately integrating numerous tasks to efficiently scale ML pipelines.

By implementing such a smart ML pipeline, McDonald’s utilizes geospatial data and standardized analytical models to perform optimal site selections across different global markets. The adoption of these technologies provides McDonald's with a significant strategic advantage on the global business stage.

Effective scaling of this process is crucial for any company. It enables expansion into new markets, maximizes efficiency in existing markets, and consequently improves customer service. McDonald’s efforts in this domain stand out as an excellent example of using ML technologies to solve practical business challenges.

Practical Considerations and Model Validation

This section discussed in detail the methodology McDonald’s incorporates in the decision-making process for selecting new restaurant locations.

A specific example effectively used in this process is Databricks. Databricks has been rated as a user-friendly enterprise-level data integration and analytics platform for McDonald’s data science team.

1. Adoption and Initial Experiences with Databricks

An essential step highlighted in this session was the enterprise-level demonstration of Databricks. This demonstration secured the necessary high-level organizational support to advance the project.

2. Efforts Towards Standardization

The importance of standardized processes focusing on further integrating data and enhancing analytics using Databricks was discussed. A unified approach to data management and analytics significantly improves the accuracy of new restaurant site selections.

3. Model Development and Improvement

Related to points mentioned by Kirk, a session featured on model development and continuous improvement. This interactive approach is key in providing data solutions tailored to actual business needs.

From these points, it is clear how McDonald’s effectively utilizes Databricks and ML models to optimize restaurant site selection. Additionally, the intertwining of data science methodologies with strategic decisions within the organization can be inferred from practical considerations and model validation.

The recent session "How McDonald’s Uses ML to Optimize Restaurant Site Selection" generated an interesting discussion on how McDonald’s leverages machine learning (ML) to identify unmet demands and optimize site selection through collaboration. Here, key points from that discussion are presented:

Identifying Unmet Needs and the Process of Collaboration

The development team at McDonald's pays close attention to first-year data and actual sales from each market when selecting locations. This approach includes considering areas with initially high sales as potential expansion zones and gradually diminishing the importance of that data over time.

Initially proposed methods include using more open-source software and data to supplement insufficient data in specific regions. Efforts to streamline the acquisition and standardization of geospatial data on a scale are underway, which is crucial for maintaining consistency across different markets.

Collaboration emerges as a vital element in addressing data challenges. McDonald's embraces active exchange of insights and solutions, which helps effectively adapt strategies to changing market conditions.

The Potential of a Collaborative Corporate Culture

McDonald’s benefits from a highly cooperative culture, positively impacting these projects. Aligning platforms and methodologies for different markets requires deviating from standard ML approaches, but open discussions greatly contribute to understanding the diverse needs and quality of data across markets.

Open cooperation between companies creates new opportunities and significantly enhances problem-solving capabilities. McDonald’s openness to incorporating ideas from various external sources plays a crucial role in its sustained growth and adaptability.


This session provided insightful revelations about McDonald’s strategic and technical framework for selecting new restaurant sites. By systematically integrating data through machine learning and fostering an open, collaborative corporate culture, McDonald’s demonstrates an effective example of business expansion and innovation strategy. These discussions highlight the importance of merging technology and cooperation for successful business execution, emphasizing the critical role of data and AI in ongoing strategic corporate efforts.

About the special site during DAIS

This year, we have prepared a special site to report on the session contents and the situation from the DAIS site! We plan to update the blog every day during DAIS, so please take a look.