Home > Paid Traffic Sources > Adult Traffic (NSFW)

Theoretical question: How many leads till statistical significance? (5)

10-20-2018 09:37 AM #1 jabong82 (Member)
Theoretical question: How many leads till statistical significance?

So right now I'm running an "Adult Dating WAP SOI" on a high volume mobile spot. I think if I ran a decent banner I could get 300 leads daily on this spot out of my funnel.

At the moment I am running 2 banners trying to figure out which one is better.

How many leads would you feel comfortable in determining a winner? (testing budget is not an issue)

Assuming one banner doesn't blow away the other in performance, how many leads each until you decide?

10 leads? 25 leads? 50 leads? More?

10-21-2018 11:04 PM #2 vortex (Senior Moderator)

Caurmen's thread should help!

https://stmforum.com/forum/showthrea...he-Winning-Ads

And you can't really use "how many leads" as a rule-of-thumb. It would depend on the difference in performance between the banners - the greater the difference in performance, the faster statistical significance will be reached. (This is assuming they're receiving the same amount of traffic of course.)

Amy

10-22-2018 01:18 AM #3 leadcloak (Member)

Originally Posted by jabong82

It's not about just leads, send same amount of traffic on campaigns and figure out their overall performance.
Amy linked to a great thread there.

LeadCloak

10-24-2018 10:09 PM #4 kingcron (Member)

I found this AB Split test calculator that shows statistical significance somewhere on STM. I use it all the time. Might help you with what you are looking for.

https://www.peakconversion.com/2012/...al-calculator/

10-26-2018 07:27 AM #5 vortex (Senior Moderator)

Originally Posted by kingcron

LOVE that calculator - thanks for mentioning it!

And here are instructions on how to use it:

https://stmforum.com/forum/showthrea...Banners-Part-1

Amy

Home > Paid Traffic Sources > Adult Traffic (NSFW)