A map's message lives in its symbology: the colours, sizes and symbols that turn attribute values into something a reader can understand at a glance. QGIS offers several styling modes, and beginners often pick at random, producing maps that confuse more than they explain. Three modes cover nearly all thematic mapping, and each one answers a different kind of question.

Use categorized symbology when your data falls into distinct groups: land-use classes, soil types, political parties, or survey responses. QGIS reads the unique values in a field and assigns each one a colour or symbol. The reader's question is 'what kind is this?', and the map answers it instantly if the colours are distinct and the legend is clear.

The trap is too many categories. A land-use map with 40 classes is unreadable, no matter how pretty the palette. Group small classes into 'Other', aim for seven or fewer colours on one map, and choose a colourblind-friendly palette rather than a rainbow. If the legend needs a magnifying glass, the map has already failed.

Colour choice is where many thematic maps quietly exclude readers. About one in twelve men has some form of colour-vision deficiency, and the classic red-green gradient is the worst offender, turning a clear map into an unreadable one for a significant audience. QGIS ships with colourblind-friendly palettes, and using them costs nothing: the map looks just as good to everyone else and becomes legible to readers it would otherwise shut out.

Test your map the way your readers will see it. Print it in greyscale and check that the classes are still distinguishable, because many maps are printed or photocopied in black and white. If two classes collapse into the same grey, add a pattern or adjust the lightness range. A map that survives greyscale printing and colourblind simulation is a map that communicates to everyone, which is the entire point.