It depends in what niche and how much traffic is "decent".If you go to adwords.google.com in Keyword planner you can see how much it costs that "decent" traffic.
You want to expirament a little bit with it to see what works. Once you get to the point where you can confidently predict that for every $X you spend, you will get $X *2 or more, then you should put in as much as you can afford. Until then, invest smaller amounts as tests.
Paid advertisement, PPC, Adwords all these need money, so I would say that test the product first, if it is giving you profit then the more you spend the better you would earn.
When you’re just starting out, you’ll want to contain your costs, because you don’t know what’s going to work. You’re in testing mode. It’s possible that your initial test campaign will be profitable, but you may only break-even, or you may lose a little money. We’ve created a lot of campaigns that were profitable right out of the gate, but you shouldn't expect this to happen. Instead, your mindset should be that you’re investing in market research.