Choosing New Restaurant Sites on Measured Foot Traffic
Taco Bell used Safari AI to measure pedestrian traffic at potential new locations before committing to a store.
Measured at each candidate site
Each new restaurant was a long bet on a short traffic study.
Site decisions often rested on a few days of manual counts or modelled data. A location that looked busy on a Tuesday afternoon could be empty at dinner, and a wrong call cost years of rent and build-out.
- Short manual counts missed evening and weekend patterns
- Modelled traffic data could not be checked on the ground
- Weak sites were found only after the store opened
Foot traffic at every candidate site, hour by hour.
Connected to site cameras via an on-premise server
Cameras covering each candidate site fed a lightweight on-premise server Safari AI deployed, so video was processed on site.
Counted pedestrians around the clock
Computer vision counted pedestrians passing each site by hour, day and week, including the dinner and late-night windows.
Compared sites side by side
Dashboards ranked candidate sites by traffic volume and pattern, so investment decisions rested on measured demand.
Note: Illustrative data
Fewer risky openings, faster decisions.
Pedestrians were counted at each candidate site by hour, day and week, including the dinner and late-night windows.
Candidate sites were ranked on measured demand, so the best location wins on evidence.
Locations unlikely to generate enough revenue were ruled out before money was committed, saving time and avoiding costly openings.
Action Center: From Insight to Action.
Measurements like these feed Safari AI's Action Center, which alerts the real estate team when a candidate site crosses its investment threshold.
- Real-time recommendations to improve KPIs via email, SMS, robocall, mobile app, Power BI, or API.
- Recommendations grounded in your own written SOPs, not generic best practices.
See this for your own site pipeline.
Taco Bell evaluated potential locations on measured pedestrian traffic. We'll walk through which candidate sites are the best fit for a pilot.
