The "Main Course" recommends setting a daily budget of $8 and spending a total of $15 before culling (of course if there's a really bad ad you might have reason to cull before hand.)
What I'm wandering is how I can adjust this type of recommendation for a really low payout.
My payout is .16 and I'm bidding I'm averaging about .04-.09 CPM. Would it be better to start cutting around $5-8 spend?
Is there a calculator that can be used to give you a good idea of a number to start culling at?
Yes if it is a very low payout, just reduce the budget.
Your budgets should generally scale with traffic source costs.
How much you spend each day is really up to you, there is no statistically grounded guide for this...
Consider the cookbook values to be for a say $0.50 payout.
If your payout is substantially lower or higher, scale your spend accordingly, provided the traffic is also cheaper.
It's common sense really - at a $0.16 payout, at $15 spend you would have spent nearly 100x the offer payout, so you should definitely have enough data to make decisions well before then.
Initially I usually only cull the variables that are eating up majority of my spend, like bad pubs or handsets/carriers. but even then I feel like $8 is way too low considering the amount of variables and traffic there are with mobile. Offers performing better/worse at different times of the day, angles, landers, aff networks, and then combining that all with apps/sites... you need to let it spend and see which areas show promise.
My approach to this is to set a $30-50/day budget and let it rip. When I hit that I see if there was anything that hogged spend and didn't convert (ads/apps:sites/lps/networks/etc), if yes I cull it, if no I check for anything that showed some promise. If there are any conversions I keep it going and increase the budget, and then from there I just rinse and repeat till I have about 3 days worth of data, then I start making decisions on the other variables to find the pockets of profitable traffic.
An example of this today with a new offer, $30 spend and only $3 in rev. There were conversions so I upped budget and culled a placement hogging about 60% of the traffic. end of the day it actually ended up in profit: https://www.dropbox.com/s/j6pm7haqdz...21.19.png?dl=0
Is it because of the time of day? is it because I culled that one site? Is it both? Idk just yet :-).
But ultimately you should have specific objectives in your testing phases, and use your budget to figure those out.
It's late and I can't sleep so I hope what I just wrote makes some kind of sense ;-)
The really key thing here is to check that you've got even coverage such that the hypothesis "this offer sucks mighty donkey balls" is more likely than any other hypothesis to explain the situation.
As such, there's no real hard and fast spend rule. Once you've spent about $5 on an offer with a tiny payout, you can certainly pause and check the data - but what you do then depends on what you find.
If all your spend has gone to two placements, it's equally likely that those placements are the ones with considerable suction in the donkey department. So you can't rule the offer out just yet.
Likewise, if one carrier ate all your spend, it might just be that the offer doesn't convert on that carrier. If a few handsets have 90% of the spend... and so on.
However, if you've spent evenly and significantly across a wide variety of placements, carriers, handsets, etc, and you've tried a couple of angles, varied banners, a lander and direct linking, then the offer starts to look like the common variable. And at that point, it might be time to shoot it in the head and give the poor donkey a break.