Step 1: determine the distance between the entry price and the stop price. Decide on how far do you want it to be away in order to avoid "false triggers". Step 2: calculate the amount of loss you'll get for one share or contract if the stop loss is triggered. Step 3: decide how much money you are willing to lose over this trade. Step 4: calculate how many shares/contracts your bet size should be. This is how you limit your loss amount per trade.
If you set your stop-loss very far away, you wouldn't be able to earn money. If you tighten your stop loss until it is extremely tight/tiny, you wouldn't be able to earn money.
One can use either recent volatility or the current price bar structure or both of those things to set stops. Using volatility means setting the stoploss back from the entry price by a multiple of the most recent ATR reading. Price structure in my case means putting the stoploss under the low of the signal bar for long entries or above the high for shorts. I've also used a combination of the two, e.g., put stoploss under low of the signal bar or N times the current ATR, whichever is farthest away. These types of rules are backtestable and can give solid results with good entries.
If you have a stop loss that actually works well, it is actually a signal to sell. Might as well just use that as an entry signal instead. Most stop lossess are bad and options are much better in withstanding the noise.