If you have seen my follow along, you notice I'm running two banners (wich I'll call here banners 16 and 19) for some days and they are doing quite well. One day they do $5 profit together... the other day $ 10 and 1 day was $40 in profit. Sometimes they break even or even lose a little bit. One thing I noticed is that from the first day I ran banners 16 and 19, they produced me a good, positive ROI (I spent like 2~3 times the payout for each on first day).
So my question is: can this be a signal to find winners? Do you have signals that some of your variations produce that tell you "I'm a Winner"
Also, what is the kind of result 1 banner + LP + Offer combo should produce you so you can tell "Yes, this variation is the winner and is ready for huge scaling"?
Still, if you had banners performing like mine, would you consider it "Campaigns Winners" or "Scalling Winners" or would keep testing to turn it into winners?
Here are the details of banners 16 and 19 + same LP + Same Offer (I'm still running it, you can see how it goes on my follow along).
There's also banner 14 here. It produced a good signal in the beggining and was good for some days. But was not a winner like 16 and 19.
Day 3 of campaign:
Banner 14 = 0.18% CTR >>> investment $ 6.58 >>> return $10 (4 conversions) <<< First day of banner 14
Day 4 of campaign:
14 = 0.16% CTR >>> 9.98 >>> 17.5 (7 conv)
Day 5 of campaign:
14 = 0.16% CTR >>> $ 20.76 >>> $ 27.50 (11 conv)
16 = 0.16% CTR >>> $ 20.08 >>> $ 32.5 (13 conv) <<< First day of banner 16. Banner 16 is a variation of 14.
Day 6 of campaign:
14 = 0.22% CTR >>> 11.90 >>> 12.5 (5 conv)
16 = 0.23% CTR >>> 9.49 >>> 12.5 (4 conv)
19 = 0.19% CTR >>> 14.46 >>> 17.50 (7 conv) <<< First day of banner 19. Banner 19 is a variation of banner 16.
Day 7 of campaign:
14 = 0.16% CTR >>> 3.19 >>> 0 <<< I stoped the banner 14 for the next day
16 = 0.16% CTR >>> 4.23 >>> 5 (2 conv)
19 = 0.20% CTR >>> 4.66 >>> 7.5 (3 conv)
Day 8 of campaign:
16 = 0.23%CTR >>> 3.58 >>> 7.50 (3 conv)
19 = 0.18% CTR >>> 3.59 >>> 2.50 (1 conv)
Day 9 of campaign:
16 = 0.24% CTR >>> 11.16 >>> 25 (10 conv)
19 = 0.21% CTR >>> 11.14 >>> 40 (16 conv)
Day 10 of campaign (Yesterday):
16 = 0.18% CTR >>> 9.53 >>> 5 (2 conv)
19 = 0.18% CTR >>> 9.50 >>> 15 (6 conv)
Hey there
Your data is really hard to read. I think you will get a much better response if you format it well in a spreadsheet with all the relevant metrics (and not just CTR) and then post a screenshot of that.
@cmdeal
Thanks, good idea, but this campaign is just an example of what happened, and don't need a deep analysis (I'm getting this help my follow along). What is most important of the whole thread is this:
Do you have signals that some of your variations produce that tell you "Hey I'm a Winner" from beggining?
Also, what is the kind of result 1 banner + LP + Offer combo should produce you so you can tell "Yes, this variation is the Winner and is ready for huge scaling"?
Always run to statistical significance.
Run data.
Stick it in a calculator like this. http://www.usereffect.com/split-test-calculator
If it calls a winner make that your new control and start another test.
if you're going for bigger gains you need 95% significance. If you're looking for 1-2% gains then you need to higher levels of significance for certainty.
@Smaxor
I'm still struggling with these statistics but I'm going to get it and this calculator will help me a lot. Really thanks.
I would suggest reading Caurmen's posts on statistical significance.
He shows what tools to use and what data to enter exactly.
Give it a go, it will remove all the guess-work.