When to cull an offer- zero convesrions (10)
How many offers are you testing there?
How many landers for each?
The amount of money you spent in them according to the screenshot is rather low given all those possibilities. You should run around 10x offer payout per each offer, as a VERY ROUGH rule of thumb (with experience in the vertical, you will make decisions on more than just data on that one test).
Also, a bit off topic, but your lander CTR looks quite low. What spy tool are you using?
Hi manu_adefy,
Thanks for replying so quickly!
Each row is an offer, have campaigns with 3 landers set up for each. The 2nd and 3rd are still running, will let them get to $20 each before making a decision. With this approach I'm budgeting to test 3x offers a day ($60/day) until I find one worth scaling.
On your other point, this is also something I've been grappling with! The 1st in particular has an atrocious CTR, the most recent 2 (same landers but different offer called out on them) seem better, but still far from what I'd be happy with. I've been using MakeMassive as it provided access to mobile and desktop Pops, which I couldn't find on WRW/Adplexity.
Each lander -> offer combo needs its own budget to get a hint of its potential.. So 3 landers + 3 offers = 3x3 = 9x testing budget. At least that's how we do it.
And we initially test only to see if there's any potential. There's no ROI goal, just testing to determine a) potential and possibly b) best lander or offer (usually won't happen just with the initial budget assigned).
I am not too familiar with MakeMassive. I had an unpaid account and their stuff looked OK, but not as good as AdPlexity. Most of your volume will end up being from mobile pops I think, for these sort of offers. It makes sense to check out AdPlexity. WRW was awful a year ago and haven't heard of any major updates, so I'd steer away of that.
What you could potentially also try is to initially just direct link the offer and see which offer is best, and only afterwards split test landers based on the best offer.
You have Traffic Source (with all its nuances) -> Banners (in some cases, not in pops of course) -> Landers -> Offer.
You can try to go TS -> Offer directly when split testing offer only. We always do LPs in too (with direct linking as well, just in case that works best), but also put in a higher daily budget to get the data ASAP.
Look for the pockets of promise within your campaign. Is there a lander offer combo doing well? If everything across your testing is showing -90% ROI or worse then I would usually move on and do more testing on other more promised offers and landers. Keep in mind to look at your targets and see if there is anything burning through your daily budget which is holding your ROI back. There's not really much of a rule of thumb everyone has there own methods of when to cut. But always look for pockets of promise within your campain. However if I've spent 10x payout and received 0 conversions, I am deff deading that campaign. Some AMs will say they test and test and test until they can make a campaign profitable. But at your stage you need to build your confidence and not break the bank.
Some really good points have already been made. I'll try to build on that. 
Basically you don't want to spend money testing offers until you're 100% sure they're duds. On the other hand, you won't want to spend so little money on testing that you're not giving the offers a chance to shine. Everybody will find some sort of balance in the middle.
Another consideration is that even the best offer will not convert on a really bad lander. By bad I mean it may not have the right angle/hook, and/or does not display/function properly (e.g. due to errors that resulted from the ripping).
This is why I always recommend to: a)Cast a wider net by testing multiple offers and landers from the beginning, and 2)stick with proven offers and landers at first.
If you start by testing 2-3 AM-recommended offers in the same vertical+geo (preferably from more than one aff network), you have better chances of seeing at least one of those offers convert well enough to optimize the camp further.
If you test those 2-3 offers using 5-10 ripped landers that you see the most often in spy tools, there's a good chance that at least a couple of the landers will perform well.
And if you test bids from the start you can get a better picture on the profits potential of the offers. A good start would be to pick 3 bids - low, average, high - and run 10-20x payout to each camp simultaneously, and see which bid give you the highest ROI, and just use that camp to continue testing your landers and offers.
(And once you're down to the best offer and lander, you can use that best lander to test every offer you can find for that geo+vertical, to find the best one. After that it would just be a matter of cutting everything that still isn't profitable - placements and other traffic segments.)
To finally answer your question of what is considered promising enough to continue?
1)After that initial bid testing (10-20x payout to 3 camps with staggered bids), I would drill down into offer -> landers and look for green combinations (and do this for all 3 camps). If I see some green, I'd continue testing (using the camp with the highest ROI). If not, I would check if there are any major traffic segments showing promise - for example if most of the conversions were made on Android, or a big carrier. All you need to have a profitable campaign, is a profitable offer+lander combo for a major traffic segment (e.g. Android).
2)So what if you don't see any promise at this point? You have several choices:
a)Ditch the vertical+geo for now.
b)Continue to test until you can make the vertical+geo work. You may want to do this if you've heard from a reliable source that the vertical+geo is killing it. If your current offers are converting so bad that it would take lots of money to test and cut landers, then just replace them with new offers and repeat the testing. If one or more of the offers are converting OK, you may want to keep running at a loss just to cut down landers - because once you have a winning lander you can use that to mass-test offers, and perhaps find a winning offer!
To cut offers, the tool you need is described in this thread:
http://stmforum.com/forum/showthread...ou-pick-offers
To cut landers here's the method:
http://stmforum.com/forum/showthread...Banners-Part-1
Sorry if I'm still being vague in my answer - would have been nice if I could provide some sort of cut-off ROI but every situation is different. But hopefully I've pointed out some of the considerations that are involved in making these decisions.
Amy
Originally Posted by vortex
Originally Posted by jurtzi
If this is the case, then the more offers you test, the better your chances of finding a good one! Just know that the more offers you test, the bigger test budget you'd need.
It really depends on what kind of test approach you want to go with. You could test 2-3 AM-recommended offers, and if you don't see promise, just give up on that geo; or you could test one geo thoroughly by testing lots of offers - either from the very start, or after you've found a proven lander.
Amy
Thanks Amy, that's extremely helpful!
I think for me, the most important thing is to get some signal in the data, i.e. conversions. I've run a couple of campaigns that have seen 0-1 conversions- obviously not enough to make any kind of decision on!
Originally Posted by rhm1991
One of the biggest hurdles faced by the newbie, is the lack of positive expectation - you’re not used to seeing green yet, or even seeing conversions.
Try this:
1)Ask for the top offers list from several affiliate networks. The more the better. Join more networks if you need to.
2)See if you can find a vertical+geo that is on many of the lists. Compile a list of all the offers you can find on all the networks you’re with.
3)Rip a bunch of relevant landers and test (for example using the approach described in the post above).
When you cast a wide net like that, chances are you will see at least one offer convert decently. Once you get used to seeing conversions, positive expectation will form and you will be more motivated to launch more campaigns.
Good luck!
Amy
Thanks Amy- funnily enough based on your advice across the forum- which is GOLD, this is actually the approach I've started taking!
I have a campaign collecting data at the moment and once the initial budget has been spent will kick off a follow-along.. but there ARE conversions
Home > Paid Traffic Sources > POP / PPV / Redirect