Hackaday

A researcher named Lauren Leek investigated how Google Maps ranks restaurants in London, revealing that its algorithm — based on relevance, proximity, and 'prominence' — creates a vicious cycle where new or lesser-known restaurants struggle to gain visibility without reviews, while chains and high-traffic venues dominate. Leek also uncovered undisclosed paid placements in Maps results. Using a machine learning model, she built a public dashboard to surface hidden gem restaurants that punch above their Google Maps weight. The analysis also revealed that restaurant diversity reflects deeper socioeconomic patterns around where communities settled and which areas remained affordable.

Google Maps Killed The Restaurant Star

<p>That’s the case with absolutely everything on the internet these days: apps, music, video… There’s a vicious cycle that keeps “average” content highly visible and forces excellent, or even groundbreaking content to invest heavily in marketing just to have a chance of breaking through… It’s all just a business thing.</p>


<p>Scoring restaurants with a model and comparing against the expected Maps rank is a clever way to surface the bias, because the residual between predicted and actual rank is where the platform’s thumb on the scale shows up. The method generalizes to any ranked marketplace. Did the gap point at review volume, or at something more like paid placement?</p>
