Campaign launching strategies for Stable Affiliates (8)
Hi guys,
I'm a bit confused on what strategy I should follow to launch my campaigns.
I already had my first profitable campaign where I did XXX profit and now I'm back to finding a campaign.
So I'm really interested in knowing what's the launching strategies for guys that are stable in this business.
1) How many campaigns per week do you launch?
2) Do you keep launching even if you find a green one?
3) If you were doing strictly mobile sweepstakes how is possible to continuously launch several campaigns?
I mean, the product is always the same either iphone or samsung, so if I launch in one GEO and several traffic sources and the campaign dies, how can I launch more campaigns in that GEO if I tried already several of these offers from different networks?
I could move to another GEO, but eventually If I was doing 2 campaigns day I would have tried in all GEOs in then what?
Maybe campaign definition is not the same for everybody. E.g. for me in mobile sweepstakes vertical I can test more than 15 offers, 5 different traffic sources, more than 10 landers, in just one GEO and I would call say that was one campaign.
If I was doing dating though, I would say each offer is a different campaign, given each offer is a different product, different landers, different type of customer.
Cheers,
Andre
Interested in this too...
Short answer: yes.
Longer answer:
You should try to saturate that vertical. Test all offers, in all geos, on all traffic sources. You have to draw the line somewhere of course, but as soon as you have an optimization and maintenance system in place, you can just launch several campaigns per day. You won't manage to saturate it of course, but you will be doing more and more volume of course.
^What manu said. 
In addition: Try to test more offers for every geo. Testing 10-20 offers in the same vertical+geo isn't as scary as it sounds. Often what would happen is you'll see a couple/few offers performing significantly better than the rest, which will allow you to cut a lot of offers pretty early on.
IMOH it's not very practical to obligated oneself to creating a set number of new camps every day. Testing widely is very important, but optimizing and scaling existing camps is just as important if not more so. Basically, keep asking yourself "what could I be doing right now that will have the most impact on increasing my total profits?" For example:
-If you have a green campaign, don't start any new camps until you've scaled it to at least every major traffic source.
-If a campaign is far from breaking even, don't spend hours analyzing placement stats to cut placements to get single-digit increase in ROI. Testing more offers to see if you can get closer to breakeven before cutting placements.
-If a bigger camp is demanding more time and attention than what you can spare, consider pausing the smaller camps to focus on the bigger one.
You get the idea. Best of luck and thanks for your question!
Amy
1) How many campaigns per week do you launch?
2 per day per vertical. We are in 4 verticals.
2) Do you keep launching even if you find a green one?
Yes, and scale the green one.
3) If you were doing strictly mobile sweepstakes how is possible to continuously launch several campaigns?
New angles
Originally Posted by vortex
In addition: Try to test more offers for every geo. Testing 10-20 offers in the same vertical+geo isn't as scary as it sounds. Often what would happen is you'll see a couple/few offers performing significantly better than the rest, which will allow you to cut a lot of offers pretty early on.
IMOH it's not very practical to obligated oneself to creating a set number of new camps every day. Testing widely is very important, but optimizing and scaling existing camps is just as important if not more so. Basically, keep asking yourself "what could I be doing right now that will have the most impact on increasing my total profits?" For example:
-If you have a green campaign, don't start any new camps until you've scaled it to at least every major traffic source.
-If a campaign is far from breaking even, don't spend hours analyzing placement stats to cut placements to get single-digit increase in ROI. Testing more offers to see if you can get closer to breakeven before cutting placements.
-If a bigger camp is demanding more time and attention than what you can spare, consider pausing the smaller camps to focus on the bigger one.
You get the idea.
Best of luck and thanks for your question!Amy
Like do you set up each offer with 5-10% in the rotator? Even if you've found an offer that's doing let's say 20% ROI today? (this offer was one of the 3 you tested initially)
Originally Posted by mj10pop
If you think about it, it's actually CHEAPER to test all of them at once (assuming you already have a decent lander you're planning to use to test all offers). Some people may want to test in batches of 5 for example, but with every new split-test, you'd be retesting the winner from the last test (as you need to use that as the "test control"), which would be money you can save.
Also, by including more offers in one test, you'll be able to cut offers sooner because the best and the worst offers are being tested simultaneously. If you test in batches, the best and worst offers may be in different "batches", so that statistical significance wouldn't be reached as quickly as if you had the best and worst offers in the same batch.
I suppose you could assign the new offers smaller rotation weight, but I don't do it that way as doing so would just drag out the testing. I want the same number of impressions to go to all test candidates during the same time period to get the testing done sooner.
Amy
I know one of my friends used to do 8k a day on facebook with that. it was only for like 2-3 weeks tho and was all small countries so they burnt out fast but hey making 150k off a few weeks is still good. Point is stick at the mobile sweeps in countries you would never think of and you might get a surprise. i still have no clue how he did it but obviously theres a way to do it
Home > Paid Traffic Sources > Mobile