The author explores how Google Maps shapes the visibility of restaurants in London, impacting their economic survival. After scraping data from over 13,000 restaurants, they built a machine-learning model to analyze how the platform's rating system works. Google Maps isn't just a directory; it actively organizes information based on relevance, distance, and prominence. Prominence is influenced by factors like review volume and brand recognition, creating a feedback loop where visibility leads to more foot traffic, which in turn generates more reviews and further visibility. This setup favors established chains and popular locations while making it difficult for new or independent restaurants to break through. To understand the true performance of restaurants beyond Google’s ratings, the author developed a machine-learning model that predicts expected ratings based on structural characteristics like cuisine type, price level, and location. They identified a “rating residual,” which indicates whether a restaurant performs better or worse than the algorithm suggests. This method reveals discrepancies between actual quality and the platform's amplification, highlighting unfair advantages in visibility. The model also accounts for issues like Google’s inconsistent cuisine classifications, further refining the accuracy of the analysis. The culmination of this project is the London Food Dashboard, a tool allowing users to search for restaurants based on various filters, including underrated options identified by the machine-learning algorithm. Although it's still in prototype form, the dashboard provides insights into London's restaurant scene, revealing how algorithmic design influences consumer choices and market dynamics. The author’s work emphasizes the need to question the assumptions behind digital platforms and their impact on local economies.