Home > Push Traffic >

---

200% ROI on the first day, I should be happy but I'm frustrated. Propeller ads (6)

---

Hey Guys,


I'm Pretty new to the AM world, I started with Amy's 40 day tutorial about two months and after trying several offers from different networks one of my AM's was really trying to convince me to run push so I had a stab at it. The offers were performing better on push but no profitable camps so I decided to give a few more offers a go. Thats when I came across this offer from NZ. I was a bit reluctant to test it as I hadn't run anything in a tier 1 geo before. The offer page looked really professional and not clickbaity so I was like what the heck, I'll give it a go. I didn't have any landing pages that would fit the offer page so I was going to run the test direct linking.


Since I didn't really know what I was doing on push I hopped on STM and started reading. I decided I was going to follow Twinaxe's push tutorial. The tutorial is great! I was happy to feel like I was understanding how things worked instead of just what to do. So with that confidence boost I started setting up my NZ campaign on Propeller ads.


So following Twinaxe's suggestions I made a few creatives that were very related to the offer such that users who were to click on them would be more likely to convert. I Opted to run Android only as I thought I can always run other OS' later if this really works. And from what I read on other threads you want to bid differently for mobile and desktop anyhow.


Onto the bid, Since NZ is tier 1 the recommended bid was higher than anything I'd seen before. Twinaxe mentioned that especially for testing, It's ok to bid lower than the average bid. that way you don't dearly overpay for weeding out bad placements and such. The recommended bid was 16 cents, and since I didn't know how much lower I could go I figured I would bid the same as I'd been bidding on other offers before that, 2 cents. I knew it wasn't only about the bid but also the CTR on my creatives. So I figured I'd work on my CTR if I wasn't getting traffic, and if anything raise the bid as a last resort. As for targeting I ran the campaign RON overnight.


The next morning when I woke up I noticed that the stats on Voluum didn’t seem to make sense. Turns out I had made a mistake and selected Pop ads instead of propeller ads as the traffic source so the tokens in my campaign url weren’t right. I quickly checked my aff network and I had three conversions. The payout is 80 cents, so I had made 2.40 in revenue. I went Propeller ads and noticed the cost had only been something like 60 cents. I was stoked, I quickly duplicated the campaign on Voluum and made the necessary changes, and duplicated it on Propeller ads. I was ready to go! At least so I thought.


Below you can see my results of the past days on the campaign. One of the days the spend is rather low because I paused it to test variations of this campaign (more on that later). The late conversions still kept trickling in though.
​


Once I started running the new campaign I noticed the offer was converting really well but the traffic volume was not enough to gather data or make significant profits, I needed more traffic. So I set out to make some better creatives. Its hard to pick a winning creative when you don’t have significant numbers. Nonetheless some were converting better than others. I looked at my best performing creatives and tried to isolate what made them the best performers and incorporated that into my new creatives. I swapped the worst performing creatives with the new ones. I was after CTR here so I made the new creatives a bit more clickbaity yet still related to the offer so as not to pay for clicks of users uninterested in my offer.


The CTR went up by a little bit but traffic wasn’t going up. So I kept trying to make better creatives. Since Propeller only allows me to have 8 creatives I ended up having to edit creatives that had already been approved. This means my CTR per creative ID is all messed up now as I’ve used some creative ID’s for two or even three different creatives. I wish I had all these stats to show you guys, I’m sure it would come in handy. If any of you know a better way to deal with this please do let me know. I thought about duplicating the campaign however I read in a thread where Twinaxe says he doesn't duplicate his campaigns during testing so I figured it was probably best not to. What I can tell you is that the overall CTR of the campaign is at 0.68%. The best creatives I made were at around 1% but since I'm always testing new creatives the campaign CTR never got there. Part of why I kept testing so many creatives was that whenever I added new creatives to my campaign Propeller would do its optimisation thing on CPM and that would bring a fresh stream of traffic. after that traffic got very slow again.


While playing around with the creatives and noticing that the traffic volume was not going up I decided to start upping the bids. I went from 2 cents to 3 cents to 5 and then to 7 cents. At 5 cents the campaign was getting a little bit more traffic(could also have been the creatives). At 7 cents the campaign was no longer profitable and volume had not gone up by a significant amount compared to 5 cents. SO eventually I left the bid at 5 cents.


But still I wanted to know what it would take to get the traffic volume I was after. I decided I was going to duplicate the campaign, pause the main camp and try to bid 14cents. Maybe that would open up some better placements and I could still be profitable? Not so much, after spending $10 and -77% ROI I stopped the campaign.The amount of Impressions Propeller was sending was allot more. Unfortunately raising the bid wasn’t going to be a viable option as my camp simply wouldn’t be profitable.


​


So I thought maybe I should be looking at other levels of user activity. That might be the way to get more volume. So I did the same thing as above, Duplicated my campaign and paused the main one. Instead of high user activity I chose medium. But this time I wanted my CTR to be the best it could be so I only chose to include my best two creatives that were close to 1% CTR. I kept the bid at 5 cents and let it rip.


​


Unfortunately my CTR didn't get to 1%, I guess it's because the quality of traffic was now worse. To my surprise the volume wasn't much higher either and the camp with medium activity was not profitable. In both duplicate campaigns I was getting a much lower conversion rate as well.


While running these duplicates I paused the main campaign as I was concerned if I let them both run they would compete against each other. Is this something I need to worry about with CPC or does it not matter since I'm only paying for clicks anyhow? I know from POP that if you want to run multiple offers in one GEO you should place them all in one campaign so that they don't compete against each other when bidding. In this thread they seem to imply that running camps side by side is ok on CPC.




Today the main campaign has been performing worse than any of the previous days. In the morning I read this thread about low volume in propeller, and how the best thing is to duplicate the campaign.
I gave in and decided I was going to duplicate my campaign. But I didnt want all my stats going to a different campaign in Voluum so I kept that the same, just duplicated on propeller. I hope that's okay.
As a result of duplicating I did get a decent amount of traffic, about 50 clicks, however the conversion rate is down. I read that it’s pretty normal to have swings in push. I’m hoping some late conversions will come in and compensate. However this has made me think twice about all the testing like crazy I’ve done.


When I was setting up the campaign I used TwinAxe’s formula for budget allocation


Average Payout * Number of Offers* Number of Landers*10


Since I wasn't using landers I just substituted it for number of creatives. Do correct me if that is wrong to do! Anyhow, payout was $0.80 I had one offer and was testing 5 creatives so I came up with a $40 test budget.


As I understand it normally you create a test budget to see if an offer has potential. you're usually looking for offers that perform at around -40% or better that you can then optimise into profit. If the offer isn't performing good enough after small optimisations during the test budget its time to move onto another offer. Since my offer was profitable straight off the bat I kind of threw that test budget out of the window.
I wasn’t really sure whether this offer that was already profitable still needed to be tested and optimised or if I should be focusing on scaling right away.


Since I wasn't getting nearly enough data to optimise I spent quite a bit of money trying to figure out how to get more volume. None of it has worked out so far and I’m concerned I may have invested more time and money then I should have.


As for placements I wanted to eliminate them after 2x payout and no conversions. But since traffic has been coming in so slow I decided to eliminate after 1x payout. I’ve only managed to eliminate 4 placements so far. What do you guys recommend? how many X payout do you use? or should I be looking at a different rule all together?


The frustrating part is that I’ve finally found an offer that converts well. With the exception of today I was getting 1 conversion every 10/12 clicks so around a 10% conversion rate. However in order to get the kind of volume I'm after I need to either have a crazy high bid or keep working on better creatives.


Previous to this offer I ran two other offers on push, both in Malaysia. Traffic was not a problem there, my CTR was 0.8% and 1.2% however the ROI after testing was -60% so I didn't pursue it any further.


I just wanted to get your opinions on whether this offer is worth pursuing. Like I said before I’m new to AM and need to work on my structure. It’s been hard to tell whether certain changes have an effect since most conversions come in much later than when the changes are made. I suppose having more volume will give me more data to work with and make things less confusing. Its very hard when 1 conversion makes all the difference since there is so little data. I've read guys like Twinaxe like to test fast. How do you do this? when moving fast how do you isolate late conversions, in other words how do you know conversions are coming in from your changes and not from things you did before.


Any tips or criticisms you may have are welcome.
If there’s any other data I can provide please let me know. As for If I do decide to pursue this offer further I thought I could go back to a low bid, say 2 cents and try to make the creatives really clickbaity while maintaining them just a bit related to the offer. hopefully that can make the CTR worthwhile for propeller. let me know what you guys think.

PS. Apologies for the long post, since I'm feeling stuck with this offer I tried to fill in as many details as I thought would be helpful (, also since I'm out of moves it makes me feel like I'm doing something useful lol). If you've made it till the end I salute you!

---

Hey @brunno, welcome to the forum.
First of all - congrats! 200% on the first day is crazy.


Different networks drive traffic in different methods, I will share my experience with our platform and I believe some of it may be useful on other networks too.
Regarding bidding under the recommended bid - Although it's fine, I usually won't recommend advertisers to do so.
The reason is that (especially in tier 1 countries) the competition very very big, people will out-bid you without even trying. What would probably happen (and I believe happened in this case) is that even if you'll be able to win some traffic and to get some conversion the offer will dry out very quickly, It's nearly impossible to make optimizations in these volumes nor to scale up.


Regarding duplicating campaigns - what we recommend our advertisers is to start with one campaign with 3-5 creatives, after having at least 1~2 profitable creatives, I recommend duplicating the campaign for each creative (meaning, one creative for each campaign), and to keep the first campaign as a test campaign - constantly testing and duplicating the successful creatives.


Regarding finding an offer that finally converts - don't let it frustrate you! Finding profitable offers is hard and it's constant work, the most experienced affiliates find it hard to find offers as well.
I recommend you to try tier 2 countries with big volumes (for example - Brazil), I know that usually payouts for tier 2 countries are lower, but the CPC is lower as well and the volumes will be bigger, I think it will be a great field for you to learn a little bit more about push-ads.


Keep your head high! Good luck.

---

Props for making it this far, Bruno. You've already taken some important steps from my perspective. First of all, if you have browsed enough around the forums you probably noticed that the biggest issues affiliates are encountering nowadays is actually finding offers that convert - at all. Then you have the offers that convert but are very difficult to maintain on a positive ROI and to optimize. And then there are offers like the one you found which actually shows some promise, so what I would do is:

1. Test it on more traffic sources and ad types. I've seen push offers going on negative ROI and doing fantastic on banner ads for example. And vice-versa.
2. Deciding on cutting down placement at 2X payout seems like a decent decision, I'd say it depends on how much the payout is and how much data you have available.
3. Rotate your creatives, test more LPs, compare CTRs and go with the top few and test them again on other sources.
4. Have you tested a whitelist on the placements that gave you the best results with a much higher bid for the sake of testing ROI?

If something works, remember that it's not broken. Clone what works and then work on improving the clone, not the initial campaign. You shouldn't worry about creating competition for your own campaign when you are bidding low. And especially if you are not bidding on highly competitive GEOs like US. You can check this if you're using the same bid initially on the clone and checking volumes/calculating the throttle.

---

Hey there, few questions.

What exactly do you mean with it?

How did you decide to run the offer?

Was it recommended or did you run it just by gut feeling?

Such things are only assumptions.

You can have offer pages that look like crap but they convert very good and you can have really high quality looking offer pages that don´t convert at all.

I don´t want to be pedantic but there are different stages of testing and depending on the stage you set a test budget to achieve your desired goal.

That can be to find a offer that is good enough to test landers, it can be test for the best lander and it also can be to see if an offer has potential.

No, there is no rule that says to find offers that perform about -40% or better.

Of course we can say the higher the ROI during the test the better the chances to succeed.

But even an offer with -80% during the test can be worth it when the remaining parameters are good.

When I have a good working funnel I usually kill placements at maximum 1 payout in loss.

When you want to run medium activity don´t target only medium - better target high and medium then together.

That way you can receive more volume and also a bit better quality.

Yes, you are 100% on spot with it.

When you run on very low volume thn the stats often are not enough to make bulletproof decisions.

When every conversion more or less can result in completely different results it´s hard to know what to do.

The more volume you run the easier it will be because then you have much more data to work with and a few conversions here or there won´t make such a big difference in the overall performance.

It will come from alone when you keep working on your stuff.

When you run campaigns long enough you will develop your own workstyle and workflow that allows you to test faster and more effective.

For example I often test lots of stuff at once.

That way I also get data fast but when nothing from my tests work it can also be pretty expensive so I wouldn´t really recommend it for a beginner.

It´s best to just stick with general and proven ways to run campaigns and when you keep busy testing stuff so that you find your own way that works best for you.

A lot of it is experience, and that´s something you can´t buy, you can only get experience by testing and running enough campaigns yourself.

I don´t care about late conversions.

They usually make only a small percentage of all conversions so I don´t take them into consideration and rather work with the stuff that I have at the time I check my campaigns.

I often have many campaigns with same targeting running together.

Often your ads are shown to a different audience anyway and even when it´s shown to the same user then chances that the user clicks each and every time they see your ad are pretty low.

Apart from all that stuff you should really consider to test landers as well.

In most cases it can help to improve performance and it als can help towards a more stable performance.

---

Hey there @popcash,

This sounds comforting

You think it's worth trying on sources such as Zeropark which don't have the option to select user activity/freshness?

Good to know. Unfortunately traffic has been so slow I was only able to cut one placement with the 2x payout criteria. I ended up cutting three more placements at 1x payout.

I have not, again since the volume was so low I figured a whitelist of say placements with 3 conversions or more would end up being a WL with 2 or 3 placements. But if I do figure out how to get more volume than that is definitely something to consider.

I Suppose I can lower my bid and keep trying to get better creatives. That way spending won't get out of hand and hopefully at some point I'll figure out creatives with a high enough CTR to make it worth propeller's while.




@am2015 I appreciate the reply.

There's no way I can be profitable bidding 16 cents with a 80 cent payout unfortunately.

Sounds to me like you're telling me to move on and test other offers.





Hey @twinaxe, thanks for reaching out.


It was indeed recommended by an AM on a new network I signed up for. She sent me a few offers in the GEOs I had been working and also this NZ offer. All these GEOs were paid together with a certain threshold. I suppose "came across" is misleading, rather it was suggested I give it a try. Since this offer was not capped but others were I ended up choosing to test this one.

Noted. My logic was that since most users in NZ will be more tech savvy than say users in some developing country they would be more averse to "crappy" looking offers. But I guess it all comes down to testing.

Also noted. You're not being pedantic, it's exactly this kind of specific breakdown of testing I was after. Seeing as I'm just starting and am looking for my first campaigns with "potential", what kind of goals should I be setting, and how should I be be allocating my budget accordingly? Like I wrote in the OP I went by your formula I found on another FA. Just swapped out creatives for landers. Although to be honest I'm not really sure exactly what goals the testing would need to achieve so that I could confidently pursue the offer. I suppose that comes with time.

Yes, -40% was just the most amount of risk I was comfortable with. I read some other posts where members were writing things like "don't insist on offers that are duds, focus on offers with -20% -30% rather than -80% or -90%", hence my generalisation. I get it, as a beginner you want to pick your battles, and in most cases -20% will be easier to get profitable than -80%(yet another generalisation). I have a friend who used to run POP and -40% was his personal go to rule so thats also why I stuck with it.

I suppose if you do some very general testing i.e. many GEOs and only one or two are profitable than you could very well have a very negative ROI while still having promising components in the campaign. Hence why breaking down by several variables is so important? The same could be said not just for GEOs but many other variables I suppose. Just blindly following my -40% rule could mean I miss out on some great opportunities.


Good to know. I suppose since I'm not there yet my criteria for killing placements should be whether the placements are adding value (data I can use) to my campaign. However there is a part of the X payout rule that doesn't seem to sit well with me. For instance if I'm killing placements at 1x, and I have an offer with a 40 cent payout. If I'm bidding 2 cents CPC than it would mean that if 20 clicks came from one placements and didn't convert I should kill that placement. But if I now double my bid to 4 cents, suddenly it becomes okay to kill placements after 10 clicks come in without any conversion. Maybe thats why I shouldn't be messing around with the bid too much during testing?

Good to know! How should I go about doing this? On the traffic source side, should I simply include medium activity in the campaign I'm running, or duplicate the campaign and include medium activity as well. If I do duplicate the campaign, do I pause the original campaign and keep the same url so that all the data is consolidated under one campaign on my tracker? Or do I set up a separate campaign on my tracker as well and run both campaigns simultaneously to do a split test of sorts?

With that being said should I continue to pursue getting volume on this offer or move on? Since I asked my AM for offers that were performing well for other affiliates, I suppose someone somewhere figured out how to make this offer work. It's still not clear to me whether this offer has potential because it was profitable or whether the lack of volume will be too big of a limiting factor.

Will do, with most other offers I tested I included LPs. The reason for not including any in this particular offer was that I was following that same logic I mentioned before. I figured if I have such a professional looking offer and use a "crappy" looking lander, I might not get so many users to end up on the offer. Hence I figured I would just test direct linking and eventually rip a professional looking lander if needed. Which leads me to another question. Are you usually looking for such symmetry in the funnel or is it not necessary? I.e professional lander, professional offer.

Since I briefly thought this Direct linking offer was going to take off and my Adplexity mobile account was due for another billing soon I decided to cancel it before I was billed for another month. I figured I'd subscribe to Adplexity push eventually when I needed more landers. Should I follow through with this? In other words can I still find plenty of landers on Push or should I just reapply to mobile?

I read your post about how the expectations of beginners is usually flawed. I also read how many campaigns you test daily. Regardless of whether I decide to pursue this offer further or not I should be testing more offers. from now I will try to test them in batches. That way I'm less likely to get attached to an offer and can be more objective about cutting/optimising.

Once again thanks for all the replies guys! Makes me feel like I'm not in this alone. I suppose these are the struggles everyone goes through in the beginning.

---

Good comments from all. No one mentioned geo yet, so I'll chime in.

NZ is a tough nut to crack. I would not cut my teeth there. Volume is low and bids are super high.

Until you get a good hang on push, I would try out some higher volume, lower bid geos with good offers in EU.

Or if you want to stick to English and Tier 1, even US on push is manageable, and you'll have a shit ton of volume, so no worries there. Just be careful with your budgets.

---

Home > Push Traffic >