Learn SEO, copywriting, conversion rate optimization, web analytics and email marketing and you'll be good to go. Or hire people with those skills if you've got cash.
Recommended resources:
SEO:
Borrowed Authority Method
Ranking Institute v2
SEO That Works
Advanced Link Building
I Want To Rank! Link Network Training
Copywriting:
Tested Advertising Methods John Caples
Breakthrough Advertising Eugene Schwartz
On Advertising Ogilvy
How to write a good advertisement Schwab
The Ultimate Desktop Copy Coach Clayton Makepeace
Advanced Copywriting Ken Mccarthy
Conversion Rate Optimization:
ConversionXL Conversion Course
Meclabs Landing Page Optimization
Meclabs Value Proposition Development
You Should Test That
Email Marketing:
The Machine Ryan Deiss
Analytics:
Advanced Web Metrics with Google Analytics
Quote:stunning images
Using vague terms like "stunning images" doesn't help anyone. What counts as a stunning image? Well if you've read some copywriting and psychological studies then you know that men tend to direct attention faster at pictures of other men and women at pictures of other women. As for other picture types that bring good results: Pictures of the product in use, pictures of the product itself, the reward that the prospect gets from using the product, pictures depicting the pain that the prospect is experiencing, pictures featuring people being dressed in strange uniforms, pictures of babies (extra effective for female audiences), etc.
It's important to test to see what works best in your particular situation. Now don't do S***** testing. By S***** testing I mean you basically "test" one image with 10 visitors and then "test" another image with another 10 visitors. That won't work. You need more people to really know what's going on. It's not rocket science. Basically you have Picture A and Picture B. You wanna know which of the pictures is more effective in getting people to your site, and is there really a difference between them. So you've had Picture A on the traffic site for a while, which we'll call "control," and now you wanna see its effectiveness compared to Picture B. So you do "testing" to figure that one out. Switch the pics and wait for results. How much do you wait? Well there's this thing called statistical significance in math. 100% proper testing requires at least 1000 people to see the ad so we can know whether the control or the challenger wins the race. Affiliate marketers who do paid traffic like Bing or Exoclick generally wait 1 or 2 days to decide the winner. They can get 1000 impressions or more for the ad on the traffic source, but obviously they can't get 1000 visitors on the landing page, so certain concessions are required. In any case they test again and again. Maybe Picture B has won over Picture A after one or two days of testing. Now what? Picture B is the new control and they test it against Picture C, aiming for incremental improvement. They do this with other elements, such as headlines and call to action, hoping to achieve the best conversion rate.
So this brings us to the research question. Who's your market? Is it both male and female? If so, try different angles with each.
Lets say you're promoting a dating product. Your primary audience is composed of males. So feature a picture of a dude being hugged by hot chicks. The picture is depicting the reward the prospect gets from the product, and it also features another man.
What kind of angle could you use in the ad? Maybe "How a virgin college student transformed his sex life using this strange trick of the ancients." That was like off the top of my head. So what did we do here? Obviously the angle is geared toward the college student who doesn't have much experience with sex, our main target market . The term "strange trick" invokes curiosity--further, the word trick indicates that the thing offered is simple, and there's also only one trick, not dozens of them, which means that the decision process is cut short and the prospect doesn't have to work much to get what he wants. People are lazy. "Transformed his sex life" indicates the benefit that the ad is offering. Remember, do not use benefits alone, but also put a curiosity element. Virgin college student: Well, if he transformed his sex life, so can I. It must be easy then, and that's the point. People want easy, fast, magical solutions. What about the "ancients" bit? It adds credibility. The "trick" has worked since antiquity, so it must be really good.
This is where ethics comes into play: Can the product you promote really offer all of the above? If not, well don't F****** promote it. Be honest in your endeavors.
And really OP you've been on the site since 2012 and yet have to ask such a basic question? I've known about IM for less than a year but it looks like our attitude toward it is dramatically different.