5 {{ upvoteCount | shortNum }}
5 {{ upvoteCount | shortNum }}

how long to test ctr

by espresso
4 replies
Hi
I am buying some banner ads at the moment
I cant rotate banners with the company so I have designed 4 banners and plan to test each
How many impression is a good calculation

I am thinking 1,000
#ctr #long
  • Profile picture of the author Warrior X
    1,000 should give you a rough idea. Make sure you're displaying all ads
    at any given time, with your software selecting between them randomly.
    Otherwise you introduce other variables (like time of day, day of the week etc.)

    Jeremy

    Edit: Except that you say you can't. Well you have to work with what you've got.
    Signature
    #1 In WHITEBOARD VIDEOS - Great Way To Tell Your Story!
    Available Here
    {{ DiscussionBoard.errors[7681800].message }}
  • Profile picture of the author WillR
    Yes, for any type of testing I think 1,000 is a minimum you should be testing with. But 1,000 impressions isn't really that much. I would be looking for closer to a few hundred clicks to really test them. But as noted above unless you are rotating them all at the same time then it won't be a very accurate split test.
    Signature

    .


    .

    {{ DiscussionBoard.errors[7682040].message }}
  • Profile picture of the author espresso
    I cant rotate the ad
    But I have a ctr at the moment of 0.2%

    Its promoting a squeeze page 50% opt in rate no upsale

    I am going to change the advert now I want a 0,5% ctr
    {{ DiscussionBoard.errors[7683053].message }}
  • Profile picture of the author espresso
    Just a heads up to anyone who cares
    My CTR rate is up to 0.3%
    {{ DiscussionBoard.errors[7683821].message }}

Trending Topics