The overall implication is that brands should treat geographic fit as a measurable factor in market strategy, tailoring propositions to local place contexts rather than assuming uniform performance across regions.

Different brands perform better in different geographies within the same category, showing that a brand’s proposition has a geography and that performance is context-dependent, as illustrated by how two chicken chains can perform differently based on local place characteristics.

The model successfully predicted reception of new locations over about 2.5 years with 1,445 openings; warm receptions matched predictions and cold receptions did as well, but the model cannot predict store openings or sales and cannot fix poorly run locations.

Researchers used Places Insights and Population Dynamics Insights to join business data with Google Maps data and to synthesize search, mobility, and environmental data into embeddings for prediction.

The study analyzed roughly 85,000 locations across eleven major brands to map brand market fit by comparing local average ratings for 15 place types in each US postal code with brand locations.

The study leveraged Google Maps Platform Places Insights (counts) and Population Dynamics Insights (population embeddings) and included eleven QSR brands, about 85,000 outlets, and 33,791 US postal codes.

Markets with similar behaviors tend to rate brands similarly, enabling cross-state learning where models trained in one state predict outcomes in others in the same direction for all eleven brands.

Location influences ratings significantly, contributing about 20.4% to a 1–5 star rating versus 79.6% attributed to the store, across 47,570 same-brand locations in the same postal code.

Practical takeaways for brands include identifying where propositions over-deliver, describing commonalities in plain terms, and finding other locations that resemble those successful places to guide expansion and messaging.

Limitations include that the model predicts reception, not where to open new stores; it does not capture individual consumer tastes, does not establish causation or sales, and cannot compensate for bad management.

Brand market fit is the difference between a brand’s rating and local expectations, not absolute ratings, and can be measured across places to forecast how well a brand will perform in a location.