Help - Clicks, Conversions and Popunders (3)
Dear Senior members, I need your help.
I have been reading all the guides that I can find here on STM, but there is some points that I’m still really confused and don’t know what to do.
Question #1 How many clicks and conversions do I need in pop campaigns before starting to optimize a campaign?
I see Vortex and a lot of other senior’s members saying that 3000 clicks with a single conversion in POP’s are a pretty normal thing to happen.
And I do believe that it makes total sense, but what I don’t understand is how can I get enough conversions to find suitable targets and still keep the budget under the 3-5x payout line.
For me it is already pretty hard to optimize a campaign with, let’s say, a 1000 clicks and 5 conversions.
I mean, with 5 conversions it’s already hard to find a trustworthy target, simply because it will be translated in something like a few clicks and 1 conversion per target.
I will give you guys an example to illustrate what I’m saying:
So, what can I do with 1 conversion and 3000 clicks (And it does happens a lot with me).
I say all this, because sometimes I get 1 conversion after 50 clicks and another conversion for the same offer just after the next 250 clicks.
What I’m really trying to say is: How can I trust in just 1,2 or 3 conversions?
Question #2: Should I use Bayesian calculators to be cutting all the other targets as soon as possible? Or should I wait for more comprehensive data (Like 3000 clicks even if with 0 conversions)?
Question # 3: Should I use landing pages from the beginning? Or can I test just offers first (Direct linking) and then pick the top converting ones and test multiple landing pages for each? (Because if I do both at the same time, I won’t know what I’m really testing, once I won’t have a control sample, right?).
Guys, any help would be great. Thanks for reading.
Hey there!
We normally say that your statistics is more ore less relevant when you spend 3xpayout. In general it's better to be taking costs into account, you don't even need the Bayesian calculator to see, that if you've spent several payouts and have 0 conversions - the campaign is not gonna work.
And using the landing page or not - depends on the type of your offer. If it's sweeps - landing is a must. But your idea makes sense, if you are working with Pin submit for example, test the offers first and then go for the landers.
Hope this helps!
Hey budapajamas,
If you haven't already, check out the following threads:
https://stmforum.com/forum/showthrea...Banners-Part-2
https://stmforum.com/forum/showthrea...t-Cutting-Tool
the 2nd link is for a placement cutting tool that uses statistical significance to help determine which placements to keep/kill/whitelist. should help you a lot in terms of decision making! the 1st link is Vortex's awesome post about it.
regarding question 3, I'd continue testing and getting comfortable with the whole process before getting into landers.
i'm sure you'll get better tips from the senior members when they chime in but hope this helps for now!
Home > Questions and Answers > General Questions