Draw 186,000 points at city scale and you get a solid blob. Point layers get three ways out of that, all in the symbology dialog (Labels & Symbols).
Which one you want depends on the question:
- Clustering — "how many are there, roughly, around here?" Nearby points collapse into counted bubbles and separate again as you zoom in.
- HeatMap — "where is this concentrated?" Density as a soft surface, no individual points.
- Hexagon — "how does this area compare to that one?" Density as crisp, comparable cells you can click.
Pick one per layer. Turning on one turns off the others.
Clustering
Switch on "Clustering" in the "Symbol" tab. Set the zoom range you want bubbles in, and choose what the number on each bubble means — a feature count, or a sum, average, minimum, or maximum of a field.
The "Template" controls the bubble's text, and "Click" decides whether clicking one zooms in to break it apart.
If you see "Cluster tiles failed to build. Points remain visible.", your points are still on the map — only the bubbles are missing.
HeatMap
Switch the symbol type to "HeatMap". Tune the radius and intensity until the pattern reads at the zoom your audience will actually use — heatmaps look quite different zoomed in and out.
Hexagon
Switch to "Hexagon" and choose what each cell should show and how big the cells are.
For classification: "Quantiles" or "Jenks natural breaks" for skewed data, which most count data is; "Equal interval" when the value scale itself means something; "Continuous (smooth)" for a ramp instead of classes.
While the cells recompute you'll see "Updating hex aggregates...".
Full styling controls are on Style Layers.
Demo data: City of Vancouver Open Data — street trees, zoning districts, bikeways, and local area boundaries.