How many impressions are enough for a "sample"?

So I found a very marketable product, and found sites on buysellads that would work perfectly, but the payout is only $1.80 per purchase on this product, granted I'd get more random purchases because of the Amazon Affiliate cookie, and I'm driving them to my own Amazon store (more potential sales), is this payout too low to make money off of a media buy?

Also, I just want to test and see my click rates and stuff, how many impressions does it usually take to get a pretty good idea if my campaign will be successful or not? 5k impressions? 20k?

Thank you

Statisticians tend to say, that statistics starts from no less than 100 samples. IMHO 1000 quality samples should be representative for any real world case.

lancis said:

Statisticians tend to say, that statistics starts from no less than 100 samples. IMHO 1000 quality samples should be representative for any real world case.

Click to expand...

Seems like it should be enough.

But I guess you never know until you try, we'll see how it works.

lancis said:

Statisticians tend to say, that statistics starts from no less than 100 samples. IMHO 1000 quality samples should be representative for any real world case.

Click to expand...

I would second this, but you are asking about impressions and that can vary widely from niche to niche and what you are paying for that impressions. But basically you will have a better idea based on actual conversions of clicks then you would on based on CTR of ad impressions. Because the latter could be more indicative of poor quality traffic / targeting or ad copy.

Time of day can also play a rule here, you need enough data to cover the day.

When you say "sample", do you mean conversion?

wasted says in his threads to drop $100 and then evaluate, I don't think he has ever said how many impressions.

Don't do anything without understanding re-targeting. I have launched many campaigns that were wildly successful at firat but then as I started to scale I lost everything I made in the beginning. Going back and looking at what I did wrong, I did not retarget. I will on my next campaign.