What explains hypercasual CTR inconsistencies from traffic to install campaigns?

Question

Hey all, I know it's a dirty word but "Hyper Casual" CTR testing via facebook.  I'm getting very wild inconsistencies between our CTR rates between our traffic / install campaigns currently.

Same budget, our succesful CTR tests (Traffic going to a coming soon web page) are at 5-7% CTR and our CTR's from our CPI (App Install) tests are around 1-2%. Meaning that the CPI is far too high. I've checked the ad set to esnure all the targerting is the same.

Currently with this data we can't really predict wether we should go ahead with a project based on the initial CTR test. Just wondering if I'm missing something and how I need to change my technique up to get more stable data.

Thanks,

Answer 1

You are totally right, you can't directly transfer ctr data from traffic campaigns to app installs, as the audience of these 2 campaigns are totally different. One likes to click on ads, another likes to install and test new games. 

Here is the rough predictions how you can extrapolate CTR numbers:

CTR vs chance for prototype to hit CPI < 20 cents:

• 3% CTR = 30% probability
• 4% CTR = 40% probability
• 5% CTR = 45% probability
• 6% CTR = 50% probability