Cheapest and least destructive first
Look before touching: habit, lustre, colour, associated minerals, and the matrix. Then streak, then hardness, then density. Only then consider anything chemical.
Each step narrows the field, and the early steps cost the specimen nothing.
Write down the result, not the conclusion
Record what each test produced, not what you decided it meant. Scratched by steel, not hardness about 5.
Observations survive a change of mind. Conclusions do not.
Know when to stop
Many species cannot be separated without analytical equipment, and no amount of field testing will close the gap. Unknown, narrowed to three candidates is a legitimate final answer.
A confident wrong label is worse than an honest uncertain one, because it propagates.
Use the associations
Minerals occur in predictable company. A specimen from a lead-zinc deposit sitting on quartz alongside sphalerite narrows the candidates before any test is run, and the matrix is often as informative as the crystal.
This is where locality earns its place in the routine. Knowing where a specimen came from eliminates more possibilities in a second than an afternoon of testing will, which is another reason an unlabelled specimen is so much harder to work with.