Home > POP / PPV / Redirect > Follow-along Campaigns

---

Levenon's POP newbie follow along (24)

---

Hi guys!

My name is Alex, 25, and I'm from Stockholm, Sweden

I have been working in Sales 9-5 since i was 16 so I have always had some sort of passion for Sales but also making comissions .
I few months back I was introduced to affiliate marketing by a friend of mine who i hadnt talked to in a long while, who has since become a big help, guide and inspiration to my goals.

However, I've decided to make this follow along because STM has been a huge help, been reading so much good stuff on here and feel like this follow along will be a motivator for me to keep going, and hopefully get some valuable tips along the way.

I've been doing POP, mainly sweeps/vouchers, focusing on SOI to learn.

Anyway, I will be updating this post for new campaigns and any help to my success would be GREATLY appreciated, and hopefully someone as new as me can learn from this aswell!


Tracking:
BeMob

Spying:
Adplexity (mobile)

Hosting:
Digitalocean (cdn at cloudflare

Affiliate Networks:
Clickdealer

Traffic Source:
PropellerAds
PopAds

---

Hey Alex,

Good idea to do a follow along.

Feel free to post as many info as possible about your campaigns so that we can try to get you rolling ASAP

---

@levenon Good luck for your follow along, sweeps in POP ca do will if you use some good landers

---

So i started a ZA campaign on propellerads, split test 2 similar offers, with 2 landers (spinners) both of them we're DOI on WIFI but 1 click flow on Cell C, so i decided to only target Cell C at first. Ran this for a couple hours with pretty low traffic volume as you can imagine..

1 hour or so in i decided to stop running 1 of the offers because the offer page looked wierd imo, with alot of russian ads all around it and had just an bad look overall.. at this point i still I had about 1k visits per offer, pretty decent CTR but no conversions.

After a while I was at about 5k visits and 0 conversions on the remaining offer, so I decided I would try targeting WIFI aswell for a while. Considered starting a new campaign only targeting WIFI to split test them that way.. maybe would have been the better strategy?

Anyway, this made the volume of traffic go up alot so I also lowered my bidding from the original 2$ cpm to 1.5 and after a while i had one conversion, although the conversion came from Cell C

I paused the campaign for now, not sure if its worth keeping it running, from what i've gathered theres no real point in optimizing anything until I have atleast 2-3 conversions?


I also noticed that theres a pretty big difference between costs in my tracker and in Propellerads, normal?

---

First things first, ZA can be a bit tricky to start with.

Few years ago ZA was one of the best Geos for Pops.

Huge volume, good bids, easy 1 click flows and very good payouts.

Because of that nice combination probably everyone who was running Pops was running ZA at one point.

That increased the bids alo so that the traffic there got pretty expensive.
Later conversion flows got more complicated and payouts got lower.

So nowadays ZA is not that golden Geo anymore and you have to compete with people there who are running campaigns for long time and have already pretty good lists they use.

ZA can still work, it´s just not that easy anymore.

Anyway, let´s check your campaign.

Did you check the offer page yourself or from where do you know about the russian ads?
In case you checked it yourself, are you sure you have seen the real offer page?

The offer is a carrier billing offer so chances are good that you got redirected and didn´t see the actual offer page when you wanted to check it.

Btw, the two offers also accept MTN traffic.
In the past I had good success with that carrier as well.

Better don´t just change the targeting or anything else in running tests.
Then you only risk to destroy your data so that you can start from scratch.

3G and WiFi behaves quite different.
They have different volume, different bids and often also different flows which leads to different performance.

When you start a campaign with 3G traffic and later in the middle of the test just switch to WiFi then you can basically ditch your test because you mixes so different things together.

It´s much better to plan your campaign, check which landers and offers you want to test, select targeting, calculate test budget and then set up the campaign and let it run as you planned it.
That´s the best way to get clean results.

When you want to test something else there wait for the previous test to finish and test it then.
But it´s better to test it in a new campaign with only the new setting, that way you have the different results from the different settings separated and it´s much easier to compare them.

I also don´t change bids in a running campaign because that can give other results and can mess up the algorithm a bit so that the campaign then behaves different.

When I create a campaign I set a bid that I want to use and don´t touch the bid anymore.
When I then want to test a different bid I create a new campaign for it.

And it´s normal that volume goes up and bids go down when you suddenly activate WiFi to the 3G campaign.

But again, these both targetings can perform very different so it´s not good to just switch between them in a running campaign.

Right, you can only optimize when you have something to optimize.

But in your case it´s a bit harder to tell because your tests are not that meaningful because you mixed different things there together and changed different variables in a running test.

I can imagine that you probably thought "when this doesn´t seem to work I just hange it to something whewre I think it´s better. That way I can save money when I don´t keep the bad stuff running".

The truth is that exactly the opposite happens.
When you change several things randomly in a running campaign you risk that your stats become useless and that you have to start the test again = the money you already spent in the test is gone.

Yes, it´s normal.
On pop traffic you have higher clickloss compared to other traffic types.

The trafficsource will show you the cost for all traffic they send you.
Your tracker can only show the cost for the traffic it received.

So when you have 20% clickloss then your tracker will show 20% less than the trafficsource.
I don´t know if it´s possible in Voluum but in Binom I just set a % clickloss so that my stats then are calculated more accurate.

Then I see that there is VAT, do yourself a favour and get a VAT number ASAP.

Now you always pay 20% VAT additional to the traffic cost, you won´t have to pay it when you enter your VAT number there.

---

Congrats on starting a follow along Alex!

I'm with twinaxe, ZA has become a very competitive GEO. I personally found US to be way easier to deal with compared to ZA (high bids, high fraud, low payouts).

Taking a systematic approach on your campaign testing would be a better option. Always check the conversion flow, as well as targeting details. Sometimes an offer can accept both Carrier and WiFi traffic, but in such cases the conversion flow may drastically change, making it way harder to make it convert on WiFi compared to Carrier traffic. So, better split the targeting into two separate campaigns.

I noticed your landing pages had some pretty good ctr, so in case you have copied them from long lasting campaigns on spy tools and don't have any auto-redirect to the offer, then it might be good to start killing all placements with no or low ctr.
Your bid might be a bit low considering the competition to this geo, so as you progressively block low-competition/quality publishers, your volume may drop, in which case you can recover it by progressively increasing the bid.

---

Duly noted. I've mostly been trying t3 countries, so good to know. I'll put ZA in the rear mirror for now.

I tried both my click URL to and also direct link to the offer page. I wasnt able to click on anything and there we're alot of ads surrounding the page.

Yeah this makes sense. Kinda thought the better move would have been to split it into a new campaign, or just let it be. Guess I was too eager to get traffic up :P

Ill look in to this right away. Thanks

Thank you so much for your answers! Will help me alot.

Also, now looking through the optimization of the campaign just to get more familiar with it all, it all makes more sense. I have no idea what traffic came when and why..

---

Hey man!

OK cool. US tho? I've only heard that newbies shouldn't start with geos like US, Canada etc because of the high competition?

OK so would you recommend to just split test carriers and wifi into 2 different campaigns as a rule of thumb? Or do you mean only to split them if I dont get enough traffic from only running one of them?

Thanks alot for your input

---

OK guys so i decided to give the US a try, split testing 2 offers and 2 landers.
I feel like im in a situation where I just randomly open up new campaigns in different geo's and tiers, trying to get something out of it to optimize, but I always get between 0-1 conversion so I just end up starting all over.
I guess thats what its all about in my learning phase though - I've made some big mistakes as you helped me with in my previous post, which didnt really give the campaign any justice, so im working on that

So this really confuses me. Why is the recommended CPM 18,5? That wouldnt ever make sense on a 2.80$ payout would it? Should i ever give this any notice? Been getting outrageous numbers on several campaigns now...
I decided to just go for a 0,5$ cpm for testing, slightly below the recommendation





SO here are the numbers. not loosing alot, and getting alot of traffic at a high tempo.
I noticed my CTR is way lower than usual, probably because of my low bid?

Is this a good point to start killing placements? or something else? Or should i let it roll for abit

---

That´s what we are here for.
To guide you through the jungle and bring some structure in your campaigns so that you don´t start random stuff

Unluckily it´s still lots of testing, there is no way around it.

But when you test with a clean and structured approach at least you learn from it because then you can analyze it much better.

I just checked myself because I couldn´t believe ir but the recommended bid is really that high.

So it was good from you that you bid lower although $0.50 probably is a bit too low.

Then I see that you excluded few Android versions and included different browsers?

Is there a reason why you did it that way?
I will be honest with you but I personally never do such tight targeting, especially not when I have no previous stats that show me such narrowed down performance.

I also wouldn´t recommend to run Android and iOS in one campaign together.

These two OS can perform pretty dfferent and bids often are also different.

Anyway, instead of starting with US right away you can also use NZ for example as a first test ground.

There are many placements that can be used for US as well as for NZ so you could start there and test stuff with less risk to lose an arm and a leg when something goes wrong

In Tier 1 Geos it´s normal to have lower CTR compared to lower Tier Geos.

In Geos like US you can kill placements very aggressive and still have enough volume left.

But then you should still re-test these placements once you have a good converting funnel to check them for quality.

---

Recommended bids on Propeller have been really broken lately. Mine always recommends pretty close to the Max Bid every time.

If you have an naked campaign (no BL/WL) and just select the platform+OS it should recommend something close to the optimal bid on their traffic chart tool

If it doesn't, ignore it and just use something close to the optimal bid for your OS for most geos. For huge geos like US, yeah, CPM of around $1 is a good place to start. But be very careful with your Daily budget, you can get flooded with tons of traffic really quickly

---

and I thank you for it.Honestly this is just something I've picked up along the way. Older versions of Android are usually worse converters, and i thought especially for US that it would be a good strategy, because it's not like im going to get too little traffic?But you think i should go just clean slate from start pretty much? no matter from geo im in?now THIS is interesting... I've been thinking about this one, since it always seems to be a very different outcome from these two.Now, would you recommend always starting 2 different campaigns split testing IOS and android? if not, how do I know what offer to go IOS/Android?Good to know! makes sense to be more aggressive with traffic in geos that gives you alot of volumeThanks, I'll try NZ

---

Hey Jaybot!

Yeah I've been giving this much thought lately.. the bidding is a thing of complexity

My thought now is just to place my CPM bid same as the offer payout, so 1 conversion/1000 puts me at 0% ROI.

might be a good starting strategy yes?

---

Probably not, but you gotta start somewhere

There is no set formula. It all depends on the geo and the offer. If the payout is super high, you'd be willing to blow a lot more money at high bids for quality traffic just to hit profit. If the payout is low, you can burn through a lot of money on high traffic zones even at a low bid.

Shit. Even I don't understand bidding entirely. It's a lot of heuristics based off a lot of factors, so the best thing to do is try to find the average bid for your particular geo + platform and start from there.

Depending on the offer and geo, getting 1 conversion for every 1000 visits may be a bit hopeful

---

Yes, a good campaign will work without such micro targeting.

In worst case you just risk to get punished for targeting too tight what could result in lower volume for your better targeting as well.

When are in test stage your only goal is to find something that´s working.

In that stage it´s not yet about scaling or most volume or whatever.

Then you should run your campaigns as (cost) effective as possible.

And usually you have more volume, lower bids and more often also better performance on Android.

So you don´t start two seperated campaigns for Android and iOS.
Just start with Android (or if you want to include Windows Phone, Symbian and Blackberry as well) and when the campaign is successful test it on iOS as well.

---

Alright, I've been going through your responses while trying different offers and also made a jump over to MaxBounty and have been testing them out for a while

So I started a Iphone 11 split test between 2 offers and 1 Landing page..
I mean , I have been trying several landers now and this is the only one thats gotten me any conversions at all.
Most people say split test 5 offers with 10 landers and so on but, are there really that many variations i can try with iphone sweeps? different styles of spinners? I'm subbed to Adplexity and it seems most people go for spinners for these kind of offers.

ANYWAYS so i started this campaign in NZ
Payout: 2$
Target:Android

I got 1 conversion in my first 1k visits or so, which made excited. So I went to bed, woke up in the morning and had about 5k visits and another conversion!
Both conversions came from 1 of the offers, and the other offer had a much worse CTR so already here i decided to go with the winning offer and pause the other.

What i realized here is that, being such a small geo, i was barely getting any more traffic. was stuck around 5k.
Excited as I was about these 2 conversions, and the fact that i was making a profit, I decided to add IOS to the campaign which boosted visits and CTR but got no more conversions..
I also removed a some IOS/android versions and placements that had 0% CTR and was costing me money

This is right now, although you can only see 1 conversion because I messed up the postback URL with MaxBounty when first conversion came in
so 4$ revenue

AAny thoughts guys?

---

Nice work!

Getting a conversion, any conversion, in NZ is an accomplishment. Keep track of those placements and start a WL/BL now. If you can make NZ profitable, you're on a really good path. Since you are past the 1 lotto conversion, and you now have 2 from that offer. I would pause the rest and try to work on that one.

Others will chime in about statistical significance and whatnot here, but NZ doesn't give you a shit lot of volume to work with.

As you said, there is very low volume.

There are like 3 million people in NZ and 60 million sheep. That means 20 sheep for each person.

Now, if we could hook the sheep up to the internet and start selling to them, then it would be a goldmine.

Anyway, wait for twinaxe to chime in, he has made NZ work really well in the past. So I defer to his wisdom.

---

Hey levenon congrats on those conversions, NZ can be tricky and performs differently a lot from offer to offer. I was also running NZ on pops on same source and I was using spinner type of lander but my CTR was much higher than yours. Spinners tend to have much higher CTR overall than other types of landers but sometimes less conversions. Ofc that doesn't have to mean anything but just maybe something to look into later on when you collect more data for that GEO. Some spinners are more appealing than others in core there is no much difference between different variations in terms of "functionality" so to speak but if you tune them to look like offer page and put some dynamic data into copy it can perform better. I wouldn't touch placements until I'm sure they are statistically significant 1 conversion can change everything, placements are last thing you want to optimize and also you could ask your AM maybe he can fire conversion so you can have better/more accurate stats. Otherwise looks promising. I'm also interested to hear twinaxe about this.

---

Hahaha, fuck yes dude, you never know what the future holds right

Thanks for the uplifting comments, felt abit down that i didnt keep the conversions going, but I will for sure try and make it work (although optimizing is scary with this volume)

---

Hey Ivan!

Great to hear your input.

I actually did, got the answer "I don't think you can get the conversion ID"

Yeah I went in after 12k visits and removed placements that had 0% CTR and had spent like 3$. Also 1 or two OS versions

---

No, you are right.
There are not that many different styles.

In the beginning 3 or maximum 4 landers will be enough to see if it works or not.
Spinwheel, questionnaire, gift boxes...

When you succeed you can still test variations of the winning style.

Don't add iOS to a running Android campaign.

You can't really compare these two.

They mostly have different volume, different bids and different performance.

It's better to keep them separated.

When you have to exclude small stuff like OS versions to get a campaign profitable then it's mostly not worth it.

About the missing conversion, unluckily on MB you can't get the click ID afterwards.

About NZ, yes the Geo is small but I had leadgen campaigns running there with average $450-$500 revenue (about $300 or so profit) per day.

Was higher payout offers but nonetheless NZ can work.

And it's a good testground with manageable risk because of the small size.

---

US isn´t the easiest Geo to start, just saying.

Then you say this

And just two sentences later you say this

2 conversions difference is not enough to call it a winner.

Optimizing by OS versions won´t help you much, either the campaign is good or not.

I can understand your intentions but the way you are testing is not the best way.

You say you have 3 LPs and 2 offers.
Let them run for some time until one of the offers converts 2-3 times.

Then use that offer to test the different LPs to identify the best lander.
Then use that winning lander to test your available offers again.

What you are doing is testing by guesswork, it´s much better to do it with a real strategic approach.

---

How many LPs do you test?

Still 3 for the US campaign?
And how many for the TH campaign?

Did you exclude the biggest placements already when they only eat your budget but don´t show good performance?

---

Your landing page ctr on TH landers seems pretty high! Are you using any auto-redirect scripts on these landers?
If so, I'd suggest testing the same lander without auto-redirect and see how the landing page ctr changes.

Also, how is the landing page ctr distributed among US campaign placements? How many visits are you giving to each placement before killing them?

Keep present that on high volume geos, you need to quickly cut out poor/low performing placements (ideally based on lp ctr after a specific amount of visits). Then as you progressively increase bids, you'll get to better performing placements when other people are seeing success from.

---

Home > POP / PPV / Redirect > Follow-along Campaigns