The Formula
Four signals, one score
SafeScore is a weighted composite of four data dimensions. Each captures a different aspect of a restaurant's food safety track record.
Normalization
One scale for every system
Health departments score restaurants differently. Some use 0–100, others use letter grades, and some just say pass or fail. We normalize everything to a single 0–100 scale.
Sources
Where the data comes from
We pull directly from government health department APIs. No scraping, no third-party aggregators. Every record traces back to an official public source.
FAQ
Frequently asked questions
Is this data accurate?
Yes. Every data point comes directly from official government health department APIs. We don’t modify the underlying inspection records — we normalize, score, and present them.
How often is data updated?
Daily for major cities (NYC, Chicago), weekly for others (San Francisco, Seattle, NY State), and monthly for Austin. Our cron jobs run automatically and we monitor for staleness.
What does the SafeScore number mean?
SafeScore is a composite 0–100 score that weighs recent inspection results, violation severity, trends, and consistency. 80+ is Safe, 60–79 is Moderate, and below 60 is Risky.
Can a restaurant dispute their score?
SafeScore reflects publicly available health inspection records. If a restaurant believes their inspection data is incorrect, they should contact their local health department to correct the source record.
Why doesn’t my city have data?
We’re actively expanding to more jurisdictions. We prioritize cities and states with publicly accessible APIs. If your health department publishes inspection data online, we’re likely working on adding it.
Is SafeEats affiliated with any health department?
No. SafeEats is an independent platform. We aggregate publicly available data but have no affiliation with any government agency.
All data is public record
Every inspection result, violation, and score on SafeEats comes from publicly available government health department records. We don't collect private data, accept payments from restaurants, or allow scores to be modified. The data speaks for itself.
Search any restaurant