mream

One of my problems in the past has been a lack of good testing of offers. I've thrown some money at something, it "didn't work" and I moved on.

This time around, I'm all set up with P202 so I get the info, and I'm planning my tests: what to look for, how much to spend, when to make a change, etc. I'm also trying to focus on a single offer & traffic source at a time.

Right now, what I'm working on is testing a particular dating offer on POF. So if you are a POF expert I need your help. Here is where I am:

- First, I tested 100 separate images with one ad. This was to pick the best images. From that, I had about 10 images that were my best (all over 0.15%). I lost money on this one, but it's ok because I was testing.
- Next, I used those images to test 5 different versions of headline and ad copy. I actually ran that another day with more new headlines and copy. I about broke even on this one. I was ok with that, too.

The plan was to take the best images and the best copy and run them together and be profitable. So far, I can't get past break even. My conversion % varies, but is acceptable overall. It's just my CTR stays low when I keep thinking that by using what's been good in the past, it will increase.

I want to get the "better" traffic, so my CPM has been competitive. My budget is not unlimited, so if I set it for ASAP, my impressions are gone in few hours. If I set it for spread throughout the day, I get a few at the beginning of each hour but that seems to have a negative impact on conversion for some reason.

Anyone have any suggestions on:
A. Testing methodology and how to pick the best ads
B. How to manage traffic on a budget

Thanks!

williamrs

I think you're testing things properly. If you have some ads with CTR above 0.15% you have already solved this problem, since this number is pretty decent.

Now, you need to get a good conversion rate a make your campaign profitable. So you need to split-test landing pages and offers.

1. Talk to your affiliate manager and pick the top 3 or 4 converting offers for your demographics.

2. Create 3 different landing pages and rotate them with P202 (if you're using a PHP geo plugin it may bring you problems, then just drop me a PM and I'll help you)

3. Rotate the offers as well and track everything

You should have a winner in few days, but if it doesn't happen the situation is clear: you're using the wrong approach. You're offering people something that they don't want, you're not passing credibility, etc... in short, you're doing something wrong and you need to re-analyze your campaign.

About your second question, I think you're doing things right, too. Spend just as much as you can afford. If your budget is $20 spend $20 a day. Things move faster when we have a good budget, but it doesn't mean that you need to have hundreds or even thousands of dollars to test a campaign.


William