[Doctor Stochastic]

The idea goes back to Linnaeus. He grouped living things by their appearance. The groups naturally fall into hierarchical groups. For example, mammals give milk, have fur and warm blood (lots of other stuff having to do with jaw and ear bones) whereas birds have warm blood and feathers; reptiles are cold blooded, etc. Or, monocot plants (grass, bananas) have a single seed pod and parallel veins in the leaves, dicots (oaks, roses) have two part seed-pods and web-like veins. Coniferous (cedar, fir, juniper, pines) bear cones and have different internal piping from flowering plants. Of course, more information and properties can be used.

Later, as genetic data became available, similarity of genomes began to be used to compare various groups of living beings. The groupings from genetic data are in essential agreement with just looking at superficial features.

A short discussion of taxonomy is available online here.

[mlc9852]

"The groupings from genetic data are in essential agreement with just looking at superficial features."

That's what I would assume as it is the most logical way for quick identification.