How to target correctly (50OnRed Pops) (8)
Hi fellow STMers
I am very new to AM, started two -99.99% ROI Campaigns....
Since I'm new and on low budget for now, I am looking out for low paying offers around $2max.
Let's take an exampe:
I take a $2 Win an Iphone Sweepstakes offer as an example:
I made 2 different lander version: 1 quiz style and one straight: Congrats blabla -> CTA
If you test something 2x-4x the payout, is this per keyword or per lander?
I take 50onRed as an example because thats where I currently run:
If I choose iphone as an keyword target, I will see traffic coming from a lot of different domains. Should I test each domain until it hit the 2x-4x payout per lander?
If I choose www.iphone.com (i know, just an example) as a domain target - test until each lander reached 2x-4x payout?
What if competition on this target (domain or keyword) is so high, that the bid is too high and while being on proxy bid, 2x - 4x the payout is less then 1k views?
I could bid lower, but then you almost don't get any traffic...
Thanks
Hey!
You can find offers with even lower payouts that $2 - this will help your budget more as you test and learn.
And yes, this is an area of confusion for many people. There’s no golden rule to follow in terms of how much to spend etc. It varies a lot from person to person.
Personally, I have a couple of thresholds…
1) Campaign - if the campaign spends over a certain amount and doesn’t show any potential, I would cut it. Caurmen suggests in his "Mobile cookbook” guide to run a sub $1 payout campaign for $15 of traffic.
2) Placements/landers - If the offer shows some potential I move on to optimising placements and allow these to spend roughly 2x the payout before deciding to cull them. For the landers, I try to make sure I have enough data to make sure my choices are based on statistical significance - you can find calculators to help you do this.
In terms of the issue about high bid placements - there’s nothing you can do about this except try to outperform the competition. If you can’t afford to compete on a placement, you need to find somewhere else or optimise your landers etc so that you can run profitably when bidding at such high levels.
I all reality spend should be correlated with statistical relevance.
Statistical relevance has nothing to do with spend or anything like that. It solely has to do with total number of clicks and then total number of goals (conversions).
Way more important is the average conversion rate % and the click cost.
Average conversion rate % will give you how much data you need to prove a theory. If it should convert at 30% you're going to get data back rather quickly with very minimal clicks. If it converts at .3% you need 1,000's of clicks. Again no factor of money is being considered here.
Click cost is independent but once you calculate how much data you need (clicks) based on you guesstimated conversion rate. Then you can calculate what it should cost you to test your hypothesis.
Thanks for the responses 
What if I choose to target on a general keyword (for example iphone or win...) in order to see what domains come in
I think that my main issue here is the missing keyword research/bad targeting for PPV - is there maybe some guide out? Im trying to find high alexa ranking URLs that somehow could fit to the offer - but thats maybe too obvious
Thanks
Originally Posted by Smaxor
We could massively over-simplify and just look at an offer being tested on a single traffic source with multiple creatives & multiple placements. If you keep all the other factors the same & just compare different placements, one placement could convert at 10%, another placement at 1%, another placement at 0.1%, another placement at 0.01% etc, so there is a huge variation around the 'average'.
Given this massive variation and also the many other factors that affect conversions, I don't understand how it's possible to arrive at a 'average conversion rate' to base things on ...or am I missing something here?
@ fallonp - I'm with you on that one.
An average conversion rate? Is this the average over all sources?
An average conversion rate on pops might be 1% while the average conversion rate on FB might be 10%. Pops are drastically cheaper than FB....
But maybe the "average" is 5% - you'll be giving up on profitable placements on pops because you're trying to hit that 5% conversion rate.
In reality an average conversion rate would be nice to have. But in reality - you don't get given average conversion rates for each source or placement etc.
This is why CR is important, but you need to look at it in the context of the CPC, the traffic source(s), the veritcal itself...
This is the game (and art) of the biz...
Originally Posted by Smaxor
Also, there are so many factors at play that reaching significance on spend may not necessarily mean the campaign is a dud. You could always make changes within that campaign and then wait again for significance. But then you are spending valuable time and money.
I think the spending cap serves to push you on to something that is potentially more lucrative rather than leave you making endless modifications to prove, with significance, that a campaign has no value.
Just some thoughts. Interested to hear what others think.
Home > Questions and Answers > Traffic Source Questions