The classic Linnaean taxonomy is basically part of your family tree without drawing in all the branches. Linnaeus was great for the 1700s, but can be very misleading today if you try to cram all the different branches of life into arbitrary horizontal categories like: Kingdom, Phylum, Class, Order, Family, etc. All of these categories are arbitrary and the only way to make them work is through lots of "sub-" and "super-". But, it's still useful to ask why the split between two taxa is being made: What do they have in common? How are they different? Especially as it relates to us. So if I want to compare and contrast myself to a frog, I go up the tree to the place where vertebrates split into amphibians and mammals. I will share all the characteristics of other vertebrates with frogs, and I will be different from frogs in all the ways mammals are. This helps you determine who is a closer relative, and this is useful in many different ways, e.g. if you want to predict how pollution might effect humans, it is usually better to look at mammals rather than amphibians, if you test a drug on pregnant rhesus monkeys, it might act differently on human mothers because we have a different placenta.
[from Dennis O'Neil 2012]
Article on fish with legs, and the evolution of vertebrate limbs.