Where the
conclusions come from
MAKE IT HERE never presents an AI guess as an established fact. Every figure in the product carries one of four labels.
Eleven factors, one number
For a business category at a location, each factor is normalized to 0–1 from the data within the category's market radius, multiplied by its weight, and summed to a 0–100 score. Category-specific weights override these defaults (a hotel weighs visitor demand more; a coffee shop weighs morning traffic and complementary businesses more). The full factor table is shown wherever a score appears.
| Factor | Default weight |
|---|---|
| Population in market radius | 12 |
| Population growth | 12 |
| Household income | 10 |
| Traffic exposure | 12 |
| Competitive headroom | 18 |
| Complementary businesses | 12 |
| Residential density | 8 |
| Daytime / employment density | 5 |
| Visitor demand | 3 |
| Nearby institutions | 3 |
| Development pipeline | 5 |
Competitive headroom compares existing competitors in radius to the number the population could support (residents ÷ typical residents per unit, calibrated from comparable-city benchmarks). Distances use great-circle math on point coordinates.
Pieces present, pieces missing
Cluster archetypes (Neighborhood Social Hub, Family Activity, Wellness, …) are defined as sets of category components observed recurring in successful districts of comparable cities. For each Medford district we count which components already exist and report completeness and the missing pieces. Archetype definitions are analyst-curated and labelled AI/interpretation; the counts are calculated.
Then → now
Cities are matched on metro population, growth, geography, income, tourism, regional role, highway connectivity, airport access, college presence, healthcare employment and housing pattern. District matches pair a comparable district's earlier state with a Medford district today; the sequence of what happened next is presented as evidence, and the "lessons" are labelled interpretation.
Grounded answers
The assistant receives a compact digest of the loaded city dataset — districts, counts, traffic, pipeline, scores, comparables — and is instructed to answer from it, to flag placeholder data, and to mark its own reads explicitly. Its output is always labelled AI.
Providers, connected and planned
The platform is built so any provider can be added or replaced. Every dataset records source, date, geographic level, confidence and last-updated. Current status for the Medford pilot:
- MockIllustrative dataset (retired for Medford)MAKE IT HERE · city · 2026 · confidence: low
Placeholder data used before live feeds were connected. Remaining only where noted (comparable-city narratives, cluster archetypes).
- PlannedTravel Medford visitor statisticsTravel Medford · city · 2025 · confidence: medium
Visitor volume, lodging demand, seasonality.
- PlannedRogue Valley International-Medford enplanementsAirport authority / FAA · facility · 2025 · confidence: high
Passenger volumes and airport-area trip generation.
- ConnectedTransGIS traffic counts (AADT)Oregon Department of Transportation · count site / highway segment · 2024 · confidence: high
Annual average daily traffic at state and local count sites plus state-highway flow segments, from ODOT TransGIS.
- ConnectedOpenStreetMap — comparable districtsOpenStreetMap contributors · district · 2026-09-08 · confidence: medium
Named businesses inside each comparable district today, categorized with the same rules as Medford.
- ConnectedCity zoning districtsCity of Medford GIS (LandUsePlanning_Service) · parcel / zoning polygon · 2026-11-03 · confidence: high
Official adopted zoning district polygons. Land-use permissions are MAKE IT HERE's screening reading of the Land Development Code use tables — confirm with city planning.
- ConnectedOpenStreetMap businesses & institutionsOSM contributors · point · 2026-09-08 · confidence: medium
Named shops, restaurants, services, hotels, clinics, offices and institutions inside city limits via the Overpass API. Community-maintained; coverage varies by category.
- ConnectedBuilding Permits SurveyU.S. Census Bureau · city (annual) · 2025 · confidence: high
Residential units authorized by building permits, by structure type, from the Census Building Permits Survey annual place files.
- ConnectedCurated major development projectsMAKE IT HERE research (local news) · site · 2025–2026 · confidence: medium
Hand-researched major housing, hotel, redevelopment and infrastructure projects compiled from Rogue Valley Times, Medford Alert, OPB and city records. Each record cites its report. Not exhaustive.
- ConnectedACS 5-year — comparable citiesU.S. Census Bureau · place · 2023 ACS 5-year · confidence: high
Population, income, age and 5-year growth for comparable U.S. cities; drives similarity scores.
- PlannedCounty Business PatternsU.S. Census Bureau · zip / county · 2023 · confidence: high
Establishment counts by industry for per-capita benchmarks.
- ConnectedAmerican Community Survey 5-Year (2019–2023)U.S. Census Bureau · census tract / place · 2019–2023 5-year (and 2014–2018 for growth) · confidence: high
Population, households, income, employment, housing and industry mix from the Census Bureau API; tract boundaries from TIGERweb. District figures are area-weighted overlays of tract estimates.
- PlannedQuarterly Census of Employment and WagesBureau of Labor Statistics · county · 2025 · confidence: high
Employment and wages by industry; employment growth.
- PlannedCity of Medford GISCity of Medford · parcel · 2026 · confidence: high
Zoning, parcels, urban growth boundary, planned development areas.
- PlannedPlaces / business datasetCommercial provider · point · 2026 · confidence: medium
Operating businesses with categories and hours.
- ConnectedOpportunity Score model v0.1MAKE IT HERE · district / point · 2026-11-03 · confidence: medium
Weighted 0-100 score combining demand, competition, complementarity and growth factors. Every score exposes its factor breakdown.
- ConnectedCluster detection model v0.1MAKE IT HERE · district · 2026-09-08 · confidence: medium
Matches district business mixes against named cluster archetypes and reports gaps.
- ConnectedAI interpretation (Lovable AI)MAKE IT HERE · varies · 2026 · confidence: low
Narrative explanations generated by an AI model over MAKE IT HERE data. Always labeled as interpretation, never as fact.