Spatial density and hotspot analysis
Heatmap Analysis Services
Density, distribution and spatial heatmap analysis, built on defensible parameters rather than a default radius.
Overview
A heatmap is a set of choices, not a picture
Bandwidth, cell size, colour classification and normalisation each change what a heatmap says. The same points can produce one broad hotspot or five distinct ones depending on the radius, and both maps look equally convincing. That is why an unlabelled heatmap is a decoration, not evidence.
We state every parameter, test the result against alternatives, and normalise by the right denominator. Raw counts almost always just rediscover where the population is; density per resident, per household or per competing venue is what actually informs a decision.
Where the question is “is this hotspot real?”, we go beyond visual density to statistical hotspot detection, which separates genuine clustering from the clumping you would expect by chance alone.
Sample output
The same data, mapped
A static preview rather than an embedded map SDK, so the page stays fast. Interactive maps are built on request as part of a visualization project.
Scope
Methods we use
The method is chosen to match the question and the density of the input data.
Kernel density estimation
A smooth continuous surface, with bandwidth selected and justified rather than left at a default.
Hexagonal binning
Equal-area aggregation that avoids the directional bias of square grids.
Grid aggregation
Fixed-cell counts when results must join to an existing raster or reporting grid.
Getis-Ord Gi* hotspots
Statistically significant hot and cold spots with confidence levels, not just visual intensity.
Local Moran's I clustering
Identifies high-high, low-low and outlier locations for pattern explanation.
Weighted density
Points weighted by revenue, capacity, review count or any attribute that matters more than presence.
Normalised density
Values per capita, per household or per square kilometre, so hotspots are not just population artefacts.
Temporal heatmaps
Density by hour, day or season, to separate persistent hotspots from transient ones.
Difference surfaces
Change between two periods or two categories, mapped directly instead of compared by eye.
Accessibility-weighted density
Density computed along the road network rather than in straight-line space.
Output formats
Delivered the way your stack expects
- GeoTIFF raster
- The density surface itself, at a stated cell size and CRS, for use in any GIS.
- GeoJSON / Shapefile
- Hexbin or grid polygons carrying the computed values as attributes.
- Static maps
- Print-ready PNG, SVG and PDF with legend, scale bar and parameter note.
- Interactive map
- A browser-based map with layer toggles and value inspection on hover.
- Analysis report
- Method, parameters, results and the limits of what the surface can support.
Use cases
What clients use it for
Sample dataset
What you actually receive
Sample project data. Note how Northgate looks busy on raw counts but is statistically unremarkable once density and population are accounted for.
| District | Points | Density per km² | Per 10k residents | Gi* z-score | Classification |
|---|---|---|---|---|---|
| Central | 418 | 62.4 | 18.2 | 6.41 | Hot spot · 99% |
| Harbourside | 236 | 41.8 | 21.7 | 4.88 | Hot spot · 99% |
| Northgate | 184 | 22.1 | 9.4 | 1.62 | Not significant |
| Westfield | 97 | 11.6 | 6.1 | -0.84 | Not significant |
| Eastbank | 63 | 6.2 | 3.8 | -2.71 | Cold spot · 95% |
How it works
Five steps, every project
- Step 01
Tell us what data you need
Share the points you have, or ask us to collect them, and the decision the map has to support.
- Step 02
We define the data scope
We agree the study area, the denominator for normalisation and the method to use.
- Step 03
We collect and process the data
Points are cleaned, projected to an equal-area system and weighted if required.
- Step 04
We validate the dataset
Parameters are tested for sensitivity, and hotspots are checked for statistical significance.
- Step 05
We deliver the final result
Surfaces, maps and a short report stating parameters, findings and limitations.
Quoted on the area covered, the number of periods and categories, and whether the hotspots need statistical testing. Tell us the question you are trying to settle and you get a fixed price first.
FAQ
Questions we get asked
What data do I need to provide?
At minimum, a list of locations with coordinates or addresses. Attributes such as revenue, visits, capacity or review counts let us weight the surface, which usually makes the result far more useful. If you have no point data, we can collect it first.
How do you choose the bandwidth or cell size?
From the scale of the decision and the density of the data, then tested. We produce the surface at several parameter values and check whether the conclusion holds. If it does not, the map is not strong enough evidence and we say so.
Can you make a heatmap from addresses without coordinates?
Yes. We geocode the addresses first and record a match quality per record. Poorly matched records are excluded or flagged, because a heatmap built on centroid fallbacks produces convincing hotspots that do not exist.
What is the difference between a heatmap and hotspot analysis?
A heatmap shows where values are high. Hotspot analysis tests whether that concentration is greater than would occur by chance, and returns a significance level. If you need to defend a decision, ask for the second.
Can you analyse changes over time?
Yes. We build surfaces for each period on identical parameters and map the difference directly, which is far more reliable than comparing two separately classified maps side by side.
What do I actually receive?
The raster or binned polygon layer with values as attributes, print-ready maps, optionally an interactive map, and a written note on method, parameters and limitations so anyone reviewing the work can check it.
More services
Often combined with
Street View Data
Large-scale street view imagery collection, processing and structured dataset delivery.
Learn MoreSemantic Segmentation
Pixel-level image analysis that turns imagery into counted, measured classes.
Learn MorePOI Data
Business and point-of-interest records with categories, coordinates and attributes.
Learn MoreKeep exploring
Where to go next
Related services
- GIS AnalysisSpatial analysis and geographic data processing, from overlays to network models.
- 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.
- Map VisualizationInteractive maps and data visualization built for clarity and fast loading.
Related reading
- How to Create a Heatmap from Location DataBandwidth, normalisation and colour classification each change the conclusion. A heatmap without those three stated is decoration.
- 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
Have a specific data requirement?
Tell us the geography, the fields and the cadence you need for heatmap analysis. You get a scoped plan, a sample and a fixed price before any work starts.