Bit late to the game but OP still seems to be logging in to BHW so I'll reply.
I'd actually recommend just trying out free bots because every idea has pretty much been done. I've yet to come across a truly unique paid bot that wasn't cloned as a free offering. The simplest way is download 10-20 bots and forward test them on demo accounts. If one works, move it to a small live account, if it doesn't then delete it. Basically, run your bots as if you had a trading floor. You have an "interview" or test period, then the good ones are promoted, the bad ones are fired.
If you want an advantage, you can get set files that are tested and optimized. The MQL forum comments often have them, and I've used OptSets before. Then you get a free bot that's optimized and tested by other people already vs something with claimed, but ultimately unproven, profitability. I'd still recommend forward testing a few days just to make sure it does what it should.
And remember expensive doesn't mean more profits, or even a better bot. The underlying strategies usually aren't unique, so the money is actually in optimizing (but not overoptimizing/overfitting). That's how human traders get good. You learn a basic strategy (something a free bot can offer) and then you optimize. You practice, you learn what works, what doesn't, you tweak it until it's profitable.
