Home > Paid Traffic Sources > Adult Traffic (NSFW)

Struggling to keep up profits (7)

10-07-2019 12:01 PM #1 mstgeorge (Member)
Struggling to keep up profits

Hi guys, one area I have been struggling lately is when a campaign loses steam and I am trying to make it work again.

After some rounds of optimization, I usually end up running a campaign with couple banners with the same angle but different images and/or gifs. I have set some rules about when to cut a banner, with the most important being when a banner doesn’t convert after two payouts, or has couple negative ROI days in a row or the ROI falls below a certain level.

The first step when a banner meets the above “rules” is to pause it and add more banners with the same angle and different images or gifs. Sometimes it works and campaign starts performing good again, but sometimes I will throw lots of same angle banners with different images and nothing will stick. At this point I have another “rule” to test up to 10 different image/gifs with the specific angle and if anything stands out, I consider the angle has “burned out” so I will try to use some proven images and test new angles. The problem is that when scenario No2 happens I lose most of the profits I have made till I find some good banners again.

For example, in a campaign that I am running right now I was running with a ROI of almost 60% for couple weeks and after scenario two happened I am under 10% and still haven’t found a good banner

I know there are a lot of factors that can affect a campaign and I cannot get a specific answer like do X if Y happens, but I guess my question is: Do you think i am doing it all wrong? and which parts of my strategy should I improve?

10-07-2019 05:01 PM #2 matuloo (Legendary Moderator)

I wouldn't say you're doing it wrong, I'm actually doing something similar, but I don't have any specific rules like 10x payout or 10 new images for this. I could be probably more systematic with this, but I rely on when I "feel" I tried enough By now, I know what a solid funnel should perform like, so when the performance isn't there I just pause it and start to make changes.

Some time ago, I experimented with angles a lot, but then I realized the best ones pretty much keep on repeating themselves and the actual message is often very similar. So these days, I got maybe a dozen of time proven strong angles that I simply try to refresh with minor changes and I focus on the actual images more. Same for the LPs, I don't spent too much time on angles, small changes and new images changes usually do the job.

Most of my "testing time" is spent on banners these days, followed by testing new offers or at least different offer urls. That's where I see the biggest effect on the final performance.

I would recommenced that you focus more on these two as well. Quite often, all it takes to revive a funnel that's losing performance, is to throw a new offer or offer url in rotation.

10-20-2019 06:08 PM #3 r4raaj (Member)

Most of my "testing time" is spent on banners these days, followed by testing new offers or at least different offer urls. That's where I see the biggest effect on the final performance.
Thanks Matej for the great tip .

11-20-2019 05:39 AM #4 profit-rex (Member)

Quote Originally Posted by matuloo View Post
Some time ago, I experimented with angles a lot, but then I realized the best ones pretty much keep on repeating themselves and the actual message is often very similar. So these days, I got maybe a dozen of time proven strong angles that I simply try to refresh with minor changes and I focus on the actual images more. Same for the LPs, I don't spent too much time on angles, small changes and new images changes usually do the job.

Most of my "testing time" is spent on banners these days, followed by testing new offers or at least different offer urls. That's where I see the biggest effect on the final performance.

I would recommenced that you focus more on these two as well. Quite often, all it takes to revive a funnel that's losing performance, is to throw a new offer or offer url in rotation.
Solid advice. Its the same for me as well, I test way more banners than landers. Angle changes can renew a banner quite a bit though, although I probably do it like once or twice a month. There is a terrific article by Matuloo on duplicating working banners in his blog. OP should check it.

To OP,
In general Affiliate Marketing works in periodic cycles. Your performance starts at a trough and as you scale it reaches the crest at some point, and after some time it returns back to the trough. Worst time to test new funnels, banners .etc is at the trough as you will have to rebuild everything from scratch. I highly recommend always having 4-5 test campaigns running even during your best days, so that when the inevitable downfall happens you already have other campaigns to keep stability in your campaigns. In my case, I have constant banner/lander testing almost on a weekly basis, and we have processes in place to do that at scale. Both in-house optimization software and a team to create and upload banners and landers (I am the only mediabuyer and I focus only on that except the dev team because I love software lol).


Quote Originally Posted by mstgeorge View Post
For example, in a campaign that I am running right now I was running with a ROI of almost 60% for couple weeks and after scenario two happened I am under 10% and still haven’t found a good banner
Affiliate marketing has two parts - creative and analytics. No matter how amazing and cutting edge your analytical skills are, if your creatives suck you will find it very hard to make money, and especially scale. When you are testing creatives, you are making a number of guesses. If all your guesses are wrong, you obviously wont have good ROI. And it happens A LOT no matter how much banner intelligence you have developed over the years. It takes me roughly 25 banners to find one good banner that does equally or better than the control. One some tests I dont find a winner at all. There are interesting articles on testing by Matuloo you can read on, but if you want a quick guide this is a simple strategy:

Round 1: Test 25 banners, Typically 5 banners in 5 campaigns -> Choose spots that have good prior performance, and good volume so that you dont see too many losses. Look for banners with early signs of high performance (High CTR, High CR). This typically takes 1-3 days. Identify 1-5 banners with early performance signs. If no good banners are found here, repeat Round 1 with another set of 25 banners. Once test is over, pause Round 1.

Round 2: Test the 1-5 banners we identified earlier across several campaigns. Perhaps 5 banners in 10 campaigns. This might need translations so that you can test across geos. Look for more proven signs of high performance (Mostly High CR, but also High CTR across several campaigns). Chances are that 1-2 banners from the 5 you choose are working really well. If not, go back to Round 1. This typically takes 1-3 days.

Round 3: You have 1-2 winners now, scale it massively across geos, traffic sources, spots - Both horizontally and vertically in all directions. Scaling should be fast. I scale from $1000 - $6000 per day in 5-6 hours (After my team creates and uploads banners which is a continuous process). Way you do that is by having a good campaigns structure that is easier to scale.

Then, you go back to Round 1 without waiting for the banners you have in Round 3 to die. Build campaigns again in parallel so that you dont have large periods of shitty profit. This is how I do it.

PS: Scaling fast strategy does not work in two occasions: (1) If your offer owner didnt give you green light. Run constant low volume for 1+ weeks in each offer and ask for green light from AM. Otherwise when you scale to massive numbers in PPL they will get scared and ask you to pause lol. (2) If you are doing BH. In BH you need to scale slowly and steadily - $10 -> $20 -> $40 -> $100 -> $150 -> $200 (And then scale across multiple campaigns). Also another obvious place you cant scale fast is if your cashflow sucks, but thats not a big problem if ROI is good. In general, try to scale fast wherever possible.

11-20-2019 09:25 PM #5 jabong82 (Member)

Quote Originally Posted by profit-rex View Post
Solid advice. Its the same for me as well, I test way more banners than landers. Angle changes can renew a banner quite a bit though, although I probably do it like once or twice a month. There is a terrific article by Matuloo on duplicating working banners in his blog. OP should check it.

To OP,
In general Affiliate Marketing works in periodic cycles. Your performance starts at a trough and as you scale it reaches the crest at some point, and after some time it returns back to the trough. Worst time to test new funnels, banners .etc is at the trough as you will have to rebuild everything from scratch. I highly recommend always having 4-5 test campaigns running even during your best days, so that when the inevitable downfall happens you already have other campaigns to keep stability in your campaigns. In my case, I have constant banner/lander testing almost on a weekly basis, and we have processes in place to do that at scale. Both in-house optimization software and a team to create and upload banners and landers (I am the only mediabuyer and I focus only on that except the dev team because I love software lol).




Affiliate marketing has two parts - creative and analytics. No matter how amazing and cutting edge your analytical skills are, if your creatives suck you will find it very hard to make money, and especially scale. When you are testing creatives, you are making a number of guesses. If all your guesses are wrong, you obviously wont have good ROI. And it happens A LOT no matter how much banner intelligence you have developed over the years. It takes me roughly 25 banners to find one good banner that does equally or better than the control. One some tests I dont find a winner at all. There are interesting articles on testing by Matuloo you can read on, but if you want a quick guide this is a simple strategy:

Round 1: Test 25 banners, Typically 5 banners in 5 campaigns -> Choose spots that have good prior performance, and good volume so that you dont see too many losses. Look for banners with early signs of high performance (High CTR, High CR). This typically takes 1-3 days. Identify 1-5 banners with early performance signs. If no good banners are found here, repeat Round 1 with another set of 25 banners. Once test is over, pause Round 1.

Round 2: Test the 1-5 banners we identified earlier across several campaigns. Perhaps 5 banners in 10 campaigns. This might need translations so that you can test across geos. Look for more proven signs of high performance (Mostly High CR, but also High CTR across several campaigns). Chances are that 1-2 banners from the 5 you choose are working really well. If not, go back to Round 1. This typically takes 1-3 days.

Round 3: You have 1-2 winners now, scale it massively across geos, traffic sources, spots - Both horizontally and vertically in all directions. Scaling should be fast. I scale from $1000 - $6000 per day in 5-6 hours (After my team creates and uploads banners which is a continuous process). Way you do that is by having a good campaigns structure that is easier to scale.

Then, you go back to Round 1 without waiting for the banners you have in Round 3 to die. Build campaigns again in parallel so that you dont have large periods of shitty profit. This is how I do it.

PS: Scaling fast strategy does not work in two occasions: (1) If your offer owner didnt give you green light. Run constant low volume for 1+ weeks in each offer and ask for green light from AM. Otherwise when you scale to massive numbers in PPL they will get scared and ask you to pause lol. (2) If you are doing BH. In BH you need to scale slowly and steadily - $10 -> $20 -> $40 -> $100 -> $150 -> $200 (And then scale across multiple campaigns). Also another obvious place you cant scale fast is if your cashflow sucks, but thats not a big problem if ROI is good. In general, try to scale fast wherever possible.
Cool write up.

How long do you stay on average of an offer using this strategy?

This seems more of like a short-term strategy, and do you find that you have enough "good offers" to keep cycling like this?

I think for me these days I'm struggling to find good offers.

Thanks in advance.

11-21-2019 03:03 AM #6 profit-rex (Member)

Quote Originally Posted by jabong82 View Post
Cool write up.

How long do you stay on average of an offer using this strategy?

This seems more of like a short-term strategy, and do you find that you have enough "good offers" to keep cycling like this?

I think for me these days I'm struggling to find good offers.

Thanks in advance.
I cycle banners and landers (landers are less frequently cycled). An offer can work really long for adult traffic. Everytime you renew banners, offer will perform better again. I had ran offers for over 4 months many times, and in average offers probably last maybe 2 months.

Biggest reason why offers had to be paused is either a pay decrease or advertiser wanted it paused as its not working for them. Good offers last long if you cycle creatives. Its a struggle to find good offers for me as well, it sucks when you know you have an optimized funnel in a geo but there is no offer lol. I am planning to go TES next year so that I can meet some direct advertisers.

The strategy to make offer work long are these things for me ->
(1) Send best quality traffic initially which mean use images of older girls, bid higher, and use good quality traffic sources (No RON and little sites in exoclick lol)
(2) Don't scale hard until your AM gives you green light (Easier for AM to do this if they own the offer) or until you run under $100 a day per offer for 1-2 weeks or so. This is because offer owners make money through sales, and leads mature only in 2 weeks or so.
(3) Send a variety of traffic sources/websites/subids than just send from one site. This is because you need atleast one subid to perform for the advertiser, if you only send traffic from one subid and it does not work you will be paused
(4) Also note that you should not send too little traffic like $50 dollars a day to an offer during the initial days, this is because sales happen much less than leads for advertiser, and they make majority of money from big buyers (sharks and whales). If you send only little number of leads chances are you wont be able to generate big buyers whom advertisers are looking for to allow you to scale. I know this is true from running revshare.
(5) And the obvious -> Talk to your AM daily and try to be their best friend

11-23-2019 04:09 PM #7 mstgeorge (Member)

Thank you really much @profit-rex for taking the time to write a so detailed answer, I have to reconsider my strategy I guess...

Home > Paid Traffic Sources > Adult Traffic (NSFW)