GeoGrub

GeoGrub is a lightweight MVP for planning short-term pop-up shop and food truck locations.

The app helps vendors compare possible stops, score operating fit, build a short route, and export a decision report. It is designed to be useful without API keys: a vendor can enter candidate spots from their own research, score the signals they care about, and leave with a practical operating plan.

Live demo: https://ianjkim08.github.io/geogrub/

MVP scope

• Create a named pop-up or food truck plan.
• Set market, vendor type, selling window, and weather tolerance.
• Add custom candidate locations.
• Place candidates on a real Leaflet/OpenStreetMap map.
• Fetch live weather from clicked map coordinates.
• Auto-fill Weather fit using current precipitation, wind, apparent temperature, and weather condition.
• Click map pins or ranked cards to focus a location.
• Zoom with map controls or mouse wheel.
• Pan the map by dragging.
• Score each location across event demand, access, weather fit, competition pressure, and permit readiness.
• Re-rank stops using a transparent scoring model.
• Remove weak candidates.
• Save the plan in local browser storage.
• Generate a route view with suggested selling times.
• Export a Markdown decision report.

Why this is different from a storefront finder

GeoGrub focuses on short-term vending decisions rather than permanent retail site selection. Its output is operational: where to test, when to arrive, what risks to check, and which location deserves first priority.

Scoring model

GeoGrub combines:

• Event demand
• Access
• Weather fit
• Competition pressure
• Permit readiness

Competition is inverted, so higher competition reduces the final fit score. Vendor type, selling window, and weather tolerance adjust the weights.

Live data

GeoGrub uses the National Weather Service / Weather.gov API to turn a clicked U.S. latitude/longitude into a live weather fit score. The app looks up the grid forecast for the clicked point, fetches the hourly forecast, then converts temperature, precipitation chance, wind speed, and forecast language into the Weather fit input.

Weather.gov is U.S.-focused. For non-U.S. markets, the app keeps the weather input editable so users can score conditions manually.

Run locally

Open index.html in a browser, or serve the folder:

python3 -m http.server 4177

Then visit:

http://127.0.0.1:4177/

The map uses Leaflet and OpenStreetMap tiles from public CDNs, so it needs an internet connection when opened.

Future production integrations

• Google Places, Google Maps, or Mapbox for place discovery and production-grade maps
• Eventbrite, PredictHQ, Ticketmaster, or city calendars for event demand
• Weather API for forecast-aware scoring
• City permit pages or manually curated permit datasets
• CSV import for historical sales and vendor notes
• Backend accounts for saved multi-market plans