For broad targeting, I would definitely optimize for conversions rather than website clicks. To mitigate risk, I would start with $5/per day if you don't have a big budget since I want to let these ads run for at least a week, if not more before scrapping the ad.
Here's a real world example based on an ad I started on Sept. 21st where I am selling sunglasses:
Targeting: USA only, Male only, 45+, Mobile Newsfeed only (always turn audience network OFF). Audience size is over 14,000,000. I based the targeting criteria off of data I gathered from prior ads.
Here's the most recent ad data as of September 26th:
Click through rate: 2.83%
Cost per click: $0.61
Clicks so far: 77
Sales: 4 - though note that one of those sales included 3 variations of the glasses.
Amount spent on ads: 47.17
Because of overall expenses this ads are closer to break even than I would like but I'm getting sales already which is a good sign. For now I'm going to keep them running to see how well Facebook optimizes them.
I'm also going to raise the price on the sunglasses once I get a few more sales to increase profitability. I had initially priced these glasses low to see what I would get for a conversion rate, but I have a different set of glasses selling at a higher price point so it shouldn't hurt conversions to up the margin.
The answer to your question, with regard to guidelines, really depends on your product price IMO. My guideline is to spend up to 2x the price of your product on an ad, unless the product price us upwards of $100 or more. It is a higher priced item then spend up to 50% of the product price. Those are the guidelines I tend to stick to. If you don't have any sales after spending $50 or so I would pause the ad.
One thing to keep in mind, is that your first ads for a product are, in my experience, rarely profitable. It takes some work to figure out how to sell a product profitably. With that in mind, really dig into the data you buy. Don't let your ad spend go to waste, as there is always things you can learn from that data.
Look at age, gender, device, anything and everything you can glean. People will tell you that, but take it a step further and determine why people responded that way to your ads. Then do tests to prove that hypothesis, Don't test things at random, r because people tell you to. Formulate your own tests to continue improving ad performance.
Ask: "Why are people 45 and older more interested in my sunglasses that other age groups? Maybe it is because those specific sunglasses bring back memories of their childhood." Now write your ads using that angle. You may need to do more than one test to prove it out, but the idea is to test everything intelligently.
Working smarter and harder is what makes the difference.