New online business?

I would get at least two and maybe three high quality animated banners made up. A good banner will draw under 1 click in 500 impressions or better on a general interest site and 1 in 200 or better for a specific interest site if the banner is prominently displayed. Once you have proven that they have drawing power then you can go out and test market and buy impressions at various sites that are related to your specific niche. There are thousands. The reason you only want to get two or three high quality banners(and the animation should draw the viewers attention, not distract them from reading so something that is relatively slow scrolling and not annoying is probably best) is because you can then brand yourself easier across a wide variety of sites. Once people become accustomed to seeing your banner they are more apt to click and/or return to your site.

The best way to test if a site is delivering the advertising you are looking for is do a test market. Once you have a proven click through rate on your banners(and it shouldn't vary much that is why it is such a great way to test website effectiveness) then it is a matter of simply going out and buying 5 or 10 thousand banner impressions on a site. That should cost you less than $20 and if you do this five or six times a month you can find yourself discovering some very effective advertising venues. Also note that as a new advertiser sites will often give you an introductory deal in order to draw you in so remember that 'everything is negotiable'.

You will quickly learn which sites are your stars and which sites are your pretenders. What you do is then consider the pretenders a loss and invest no more money than your $20. Those losses should be more than made up for by your stars. So let's assume you have bought five ads for $20 each. Each site gives you 10K banner ads. Site 1-3 gives you 2 clicks each for your ad and no sales. Site 4 gives you 15 clicks and 8 sales and site 5 gives you 30 clicks and five sales. Now if you know your banner gives you about 1 click in 500 and 30% sales off those clicks then you know that 4 and 5 are dealing with you relatively square and 1 - 3 are playing games with you or just aren't up to your standards. Based on that it would be prudent to not spend another penny on the first three sites and ramp up your advertising on the last two.

When the next month rolls around you have two brand new advertising sources and then you go out and buy five new $20 test ads and do the same thing. That way you can grow your ad spaces while at the same time limit your risk of loss. I have found that this method causes the good advertising sources you find to more than pay for the ones you lose money on once your machine gets rolling. What happens is you only spend one month on the bad ones but the good ones can become ad budgets of hundreds or thousands of dollars for years so you see how the good can outweigh the bad