Sample project
Restaurant Location Intelligence
Mapping a city's food and drink offer, measuring competitive density, and producing a ranked shortlist of neighbourhoods for a new venue.
- Study area
- One city, 42 neighbourhoods
- Venues collected
- 3,180 food and drink
- Catchments
- 5 / 10 / 15 min walk
- Output
- Ranked shortlist of 6
- Duration
- 4 weeks
- Venues collected
- 3,180
- Coverage recovered
- +6%
- Areas shortlisted
- 42 → 6
- End to end
- 4 weeks
Two sources, reconciled and deduplicated
Venues missing from either single source
Scored on four weighted components
Collection, analysis and recommendation
Problem
What needed answering
An independent operator planning a second venue had narrowed their options to a general sense that the east of the city felt promising. That impression came from visiting on weekends, which is a real signal but a biased sample.
They needed to know three things: how many comparable venues already operated in each candidate area, how well those incumbents performed, and whether the resident and daytime population could support another venue at their price point.
No usable dataset existed. The venue list had to be built first, and cuisine labels across sources were inconsistent enough that a naive count would have been off by a substantial margin in exactly the categories that mattered most.
Data sources
What went in
Map platform listings
Publicly listed food and drink venues with category, coordinates, rating, review count and hours.
Business directories
A second public source, used to cross-check coverage and catch venues missing from the first.
Census small-area data
Resident population, age structure and household income by small area.
Workplace population
Published employment counts, used to model daytime demand.
Pedestrian network
Open mapping street data, used to compute walk-time catchments rather than radii.
Method
How it was done
Venues were collected across the whole city using a tiled search geometry rather than keyword search, so coverage was even instead of concentrating on the centre. Two sources were collected and reconciled, which recovered around 6% of venues missing from either source alone.
Cuisine labels were mapped into a single agreed scheme, with the original source label kept in its own column. This mattered: before standardisation, three labels that all meant the same thing split one category into three apparently small ones.
Walk-time catchments of five, ten and fifteen minutes were computed on the pedestrian network for each candidate site. Straight-line radii would have overstated reachable population in areas cut by a river and a rail corridor.
Competitive density was calculated as comparable venues per 1,000 reachable residents plus workers, not as a raw count. Review counts were used to weight incumbents, on the reasoning that a venue with 2,000 reviews competes for more of the same demand than one with 30.
Each neighbourhood received a composite score from four weighted components. The weights were set with the client, and the ranking was re-run twice as they adjusted them, which is the point of building it transparently.
Processing
The pipeline
Each stage produced an artefact that the next stage consumed, so any result can be traced back to the input that created it.
- 01
POI data
GeoJSON3,180 food and drink venues collected across the city from two public sources, deduplicated on normalised name plus a 40 m distance threshold.
- 02
Competitor distribution
Weighted point layerVenues filtered to the comparable set by cuisine and price level, then weighted by published review counts as a demand proxy.
- 03
Density analysis
Density tableComparable venues per 1,000 reachable residents and workers, computed within walk-time catchments on the pedestrian network.
- 04
Heatmap
Density surfaceKernel density of the comparable set, normalised by combined resident and daytime population, to surface under-served pockets.
- 05
Location recommendation
Scorecards + mapNeighbourhoods scored on demand, competition, incumbent performance and accessibility, delivered as a ranked shortlist with per-area profiles.
Visualization
The output
- Restaurants · 34
- Cafés · 22
- Retail · 26
Data
Results table
| Neighbourhood | Comparable venues | Reachable demand | Venues per 1k | Avg rating | Score |
|---|---|---|---|---|---|
| Riverside North | 6 | 18,400 | 0.33 | 4.1 | 84.6 |
| Old Mill | 9 | 21,900 | 0.41 | 4.2 | 79.2 |
| Station Quarter | 22 | 34,600 | 0.64 | 4.5 | 71.8 |
| East Gate | 4 | 9,800 | 0.41 | 3.9 | 66.3 |
| Market Square | 31 | 29,200 | 1.06 | 4.6 | 52.4 |
Findings
The key comparison
- Riverside Northunder-served0.33
- Old Millunder-served0.41
- East Gatelow demand0.41
- Station Quarter0.64
- Market Squaresaturated1.06
Result
What the analysis showed
Market Square, the area the operator had assumed was strongest because it felt busy, had the highest competitive density in the city and the highest incumbent ratings. Entering there would have meant competing with well-established venues for demand that was already fully served.
Riverside North had a third of the competitive density with over half the reachable demand of Market Square. It scored highest overall despite feeling quieter on a weekend visit, because much of its demand is workplace-based and shows up on weekdays.
East Gate had equally low competitive density but ranked fourth, because its reachable demand was less than half that of the leading areas. Low competition on its own is not an opportunity.
The final shortlist of six neighbourhoods included two the operator had not previously considered, and excluded three they had been actively looking at.
Deliverables
What was handed over
- Full venue dataset of 3,180 records with standardised categories and the source label retained
- Category mapping table documenting every standardisation decision
- Walk-time catchment polygons at 5, 10 and 15 minutes for each candidate area
- Competitive density table by neighbourhood, with weighted and unweighted variants
- Normalised density surface as GeoTIFF plus an interactive map
- Ranked shortlist with one-page scorecards per neighbourhood and a written recommendation
Keep exploring
Where to go next
Related services
- POI DataBusiness and point-of-interest records with categories, coordinates and attributes.
- Location IntelligenceLocation-based market, catchment and competitor analysis that ends in a recommendation.
- Google Maps DataStructured business listings with coordinates, categories, ratings and opening hours.
- Heatmap AnalysisDensity, distribution and hotspot analysis that stands up to statistical scrutiny.
Related reading
- Restaurant Location AnalysisBusy does not mean available. The areas that feel best on a Saturday visit are often the ones with the least room for another venue.
- What Is Location Intelligence?Mapping shows you where things are. Location intelligence tells you which option is stronger, by how much, and for what reason.
Next step
Want something similar for your market?
Tell us the geography, the category and the decision. We will scope the equivalent project and send a fixed price before any work starts.