They isolate the chemicals they want to test for smell through extraction/synthesis/purification and then have a professional smeller identify what scents that particular chemical matches.

They can also compare the isolated chemicals to known scents containing the same chemicals.

The pads from the teenagers’ armpits had two steroids present – 5alphaandrost-16-en-3-one and 5alphaandrost-16-en-3alpha-ol – which smell of sweat, urine, musk and sandalwood. They also had higher levels of six carboxylic acids, which give off unattractive smells including cheese, goat and wax.

What this means is that two steroids (5alphaandrost-...) found in teens' armpits are also found in sweat, urine, musk, and sandalwood. The 6 carbolic acids are also found in cheese, goats, and wax.

So to answer your question: not subjective. This is objective, reproducible evidence.

We have catalogued many chemicals and what they smell or taste like. You can thank the perfume industry for a good portion of the smell based science. A lot of flavor based research exists, as well. Unfortunately some of that research was done for sick purposes like finding out how to engender craving or how to make the body ignore chemical signals that it is full and it is time to stop earing. But I digress.