If you want to jump right into ppc/affiliate offer stuff, you're most likely going to lose money in the beginning. If you test enough offers, hopefully you will find something that works and it will cover your losses for the other offers you lose on. That and ad credits for signing up with ppc search engines make it so that if you're careful, you may never go into the red. I never did.
If you ask your affiliate manager what to promote, likely they will tell you all of the saturated stuff- because that's what they are making the most money on usually.
I would test, test, test. Look at the offer landing page and pretend you're semi-retarded, overly cautious about who you give your info to online, and remotely interested in the niche/offer. Then test offers you think would convert based on how you view the offer page from that state of mind. Basically the terms need to LOOK good at first glance, the page has to be pretty and the payout needs to reflect how much work your traffic will have to do to get you paid.
The work part pretty much goes- zip, email, CC. People don't give a shit about giving their zip out for the most part because they can't be identified from it. For the most part, people realize most Zip offers are stupid, so conversions won't be as high as you'd expect. People are overly cautious about email, but most usually have a email they don't mind being spammed at and use it to sign up for things. Also, if they think the email is just to confirm something and they don't think they will get spammed, they will give it up. If you want them to bust out their credit card, they have to really trust the site and want what it's offering. On top of that, the deal needs to sound EXTRA sweet.... then you can even count on less people converting, so make sure the payout is nice.
If you're doing free trials or other items where the user just pays shipping, and you're using PPC so you can target your traffic however you want- go for higher payout stuff. If people have to pull out their credit card, they don't care if it's $1.99 or $4.99 for shipping. Some of those offers will be like $16 payouts and others $75. You usually have a better chance at profiting with the higher payouts since they must be convinced to give up their CC info online.
Other offers that work well are things like PayDay loans or quotes for mortgages and insurance and shit like that. The company makes so much money per conversion that they can pay a lot per lead. If someone gets a payday loan, the interest rate is outrageous. On mortgage loans, they make hundreds of thousands over the years from one loan. The leads are usually not that difficult to come by because you really only need to get the person to get a NO OBLIGATION quote. The only problem with these is that they are saturated with competition, so it's not as easy as it used to be.
Once you've picked an offer or multiple offers, build a landing page. Be creative and try to build a page that would be a first step in the offer- but also sells the offer. I usually just recreate the same template the offer page uses and add my content. If you want to up your QS for Google, throw up a few pages and link them all to each other on the bottom. Research the niche and write up some content related to it. I usually put this below the template, but if it's real good, I make it a prominent part of the page. Use flashing arrows or something similar to grab your user's eye and point them directly at some real catchy sales text that they can't refuse. Have that set them up for the offer and lead them to it. Depending on the offer, landing pages vary a lot- so just search for a niche and check out some of the ads landing pages. Don't click on the ad though- naughty naughty!
This is just a start and I really didn't cover much. Half of it probably didn't make sense, but I have to run so I can't proof read.