Voronoi diagrams answer one of the oldest questions in spatial analysis: for any location, which of my points is closest? Draw the boundaries where “closest” flips from one point to another and you get a honeycomb of polygons — one per point, jointly covering the whole plane. Hydrologists call them Thiessen polygons and have used them to weight rain-gauge data for over a century; the same construction underpins service-area estimation, school catchments, retail trade areas and nearest-facility analysis.

Generating one from your own points

You don’t need PostGIS or a QGIS processing chain for this. Upload a point layer — GeoJSON, shapefile, KML, GeoPackage, or a CSV with coordinates — to this Voronoi diagram generator and it computes the polygons in your browser and exports them in any of the same formats. Each polygon carries the attributes of its seed point, so a join is unnecessary — the store’s sales figures are already on its trade-area polygon.

Two practical settings matter:

• The clipping envelope. A mathematical Voronoi diagram extends to infinity; edge polygons must be clipped to something. Clipping to the point layer’s bounding box is the default, but clipping to a real boundary (a city limit, a state) gives far more honest edge polygons — pair the generator with a boundary download tool if you don’t have one handy.
• Duplicate points. Two identical coordinates produce a degenerate diagram. Deduplicate first if your data comes from repeated surveys.

When Voronoi is the wrong tool

Voronoi polygons encode a strong assumption: straight-line distance is what matters, and every point competes equally. That breaks when travel happens on a network (drive-time service areas can look nothing like Thiessen polygons), when facilities differ in capacity or attractiveness, or when a continuous surface is what you actually want — in that last case interpolation is the right family of methods, and the Voronoi construction reappears inside one of them (see our explainer on natural neighbor interpolation).

Related constructions

The Voronoi diagram has a famous twin: connect every pair of points whose polygons share an edge and you get the Delaunay triangulation, the standard way to build a TIN from survey points — a TIN generator covers that direction. And if what you need is not polygons around existing points but a regular lattice of new ones, that’s a point grid generator instead.