Home > Mobile > Follow-along Campaigns

---

Thien's Mobile Pops Follow Along (32)

---

Hi guys, I am stating this follow along in hopes of getting advice and feedback for my mistakes. Thanks to everyone taking the time out to read this.

Geo: VN
Traffic: PopAds
Offer: iPhone 6
Payout: $0.56

Right now I have 3 campaigns set up in PopAds with all the same settings but with different bids. Targeting both wifi and carrier since the offer accepts both.

low (.001)
mid (.004)
high (.007)

I set up 3 campaigns in Voluum to track the different bids. I also have 4 landers in rotation in all 3 camps to split test. I topped up with $100 and set a budget for $33 for each camp. I am not sure if the is the correct way to do it but I am just going to go for it.

The landers are ripped from Adplexity and translated from Fiverr. I added back button redirect, entry pop, exit pop, vibration to all landers. The landers are hosted on Amazon CDN.

Right now the campaigns are pending approval, I will run them and report back with the numbers. Again any advice or criticism is much appreciated.

---

Quick update. I guess I will not be using the high bids. I started all 3 campaigns at 8 AM this morning and the high bid burned through the whole budget in half an hour. It did produce 7 conversions though but still, that was way too fast I think. I will update again when the other the camps have more numbers.

---

Another quick one. I don't think I will be using the middle bid neither because I just noticed the budget is burned through also. It did yield 12 conversions but still too quick for me to test. So I am starting to wonder am I bidding too high? So now all I have is the low bid camp running but no conversions yet. We'll see what happens.

---

Update

So here are the stats for today for testing the different bids. The high and mid bid burned through the budget way too fast. The low bid took about 8 hours to burn through the budget, the mid bid lasted 2 hours, and the high bid only lasted about 30 minutes lol. The ROI for all the bids were horrible so I think I am going to create a new camp and set the bid at .0025 and just start testing different landers. I will update again once I get more data.

---

Update

So I started the camp at 8 AM today and $100 lasted only about 5 hours. I think I may have to lower bid again because I am spending way too fast i think. Is this a good way to test or should I slow down the traffic and spread the budget throughout the day? It looks like lander2 (whatsapp lander) is the winner so I will cut the other 3 and add more landers. Unfortunately I've spent my budget for the day so I will have to wait until tomorrow to run more traffic. Should I lower my bid to spread the traffic throughout the day? If I do lower the bid, would that mean all of today's data would be skewed and irrelevant? Thanks for any advice.

---

UPDATE

Ok so I have added 3 new landers to the rotation to split test against lander2

lander2 (whatsapp lander)
lander5 (purple spinwheel)
lander6 (zalo quiz)
lander7 (coogle spinwheel)

I was not sure what to do with the bids so I kept it .0025 but I enabled the traffic throttling feature so I don't blow through my daily budget in 5 hours again. I am still unsure but by doing this did I skew my stats from the first round? Anyway I topped up another $100 and started the camp again. This time my budget lasted 26 hours before it was gone, I attached my current stats to the post. I ran the numbers through peakconversion calculator and it looks like I will need to cut lander5 and lander2. This is where I stand now, I will rip more landers to split test and report back with more data. My ROI is still horrible but the second round of lander testing improved by 2% so I guess I am headed in the right direction. Any advice is appreciated.

---

Hey Thien, great follow along. Have you tried split testing offers also? Also I would split test IOS vs Android, tablet vs smartphone.. Popads has a lot of targeting options to work with Vortex has a great article for pop traffic optimization here
Also, I usually budget 15x payout to test each lander.
I'm subscribed to this thread!

---

I've always noticed android to convert better than iphone regarding iphone sweeps. I would test that if I were you. Also, you should test more offers from multiple networks. Best of luck

---

Hey thanks for the suggestions. I did read that guide by Vortex and it is a good one. I am actively looking for more offers to split test. I am having a hard time finding offers that allow for aggressive angles. I do not plan on cloaking because I don't think I can afford it and I still have no idea what I am doing so I am holding off on cloaking for now. None of my AMs seem to want to suggest any aggressive offers. I'm not sure if that is something they are uncomfortable talking about or not. It seems the general advice is to run aggressive but I can only find non aggressive offers at the moment.

---

Hey thanks for replying, as stated above I am on the hunt for more offers.

---

UPDATE

So here are the landers for this round

lander6 (blue zalo)
lander7 (coogle spinner)
lander8 (fb survey)
lander9 (red spinner)

I topped up another $100 and started the campaign and this time it ran for about 12 hours before I noticed lander6 and lander8 needs to be cut because they both reach statistical significance(iI think?) I attached the most current stats. Now my question is should I run lander7 and lander9 against each other until one reach SS, or should I throw some more landers in the mix? My guess is more landers so I will pause for now and rip some more and update this when I get more data.

---

UPDATE

I ripped a couple more landers to split test:

lander7 (coogle spinner)
lander9 (red spinner)
lander10 (ios theme)
lander11( fb spinner)

I restarted the camp with these 4 landers and it ran for about 8 hours until I noticed lander9 reached statistical significance and beat the other 3 landers as you can see in the stats attached. Also the CTR for this lander was really high so it seems to be performing pretty well. At this point should I rip more landers or try out different variations of this lander? Anyway I'll be busy this weekend so I may have to start the camp back up either Monday or Tuesday.

---

Here are my overall numbers since the start of the campaign.

---

I would keep running lander 9, you already tried many landers and seems like lander 9 is in the green.. Now start searching for targets to blacklist, chek in Voluum the website ID's that are losing the most money. Also look at the quality score, offers I'm running in VN with popads usually convert better using quality score 6+, but chek your data in Voluum.
Vortex has a great guide on cutting placements: http://stmforum.com/forum/showthread...Banners-Part-2

---

UPDATE

I took anty's advice to continue running lander 9 so I topped up another $100 and restarted the camp. Here are the stats for the day after running it for about 24 hours. Even though I am still deep in the red, I seem to have improved my ROI by about another 10%. I will started analyzing the stats to see if I can optimize and improve my ROI even further.

---

Thanks for the advice anty. How do I check my quality score in Voluum?

---

If you have setup 'quality' as a custom variable for Popads under traffic sources in Voluum. Then you can sort by selecting the campaign and clicking on 'Report' in Voluum.

Here is the article in Popads on tracking tokens.
https://www.popads.net/kb/advertiser

---

Hey sahil thanks for the reply, I will look into this.

---

Hey good job so far! VN was one of my first geos I tested that I almost got to green. It's definitely a good one to start with.

Just wondering, are you split testing Android vs IOS? Also how many placements have you cut? With VN you probably have a LOT of placements that you can cut. Feel free to export all your data combined since you started running the camp and cut placements at a reasonable level like 2x spend or even 1.5x spend. I know there tend to be some placements that sometimes go unnoticed and eat up a lot of budget.

You might want to also consider testing more bids. 0.007 is a VERY high bid for VN. Try some more bids around your lower one. I think Popads goes all the way down to 0.0003 or something. It might even show 0 traffic for the lowest bid but it always does even if there actually is traffic at that bid. You can also test some slightly higher bids like 0.0015 and 0.002. Take a look at the bidding spectrum in the estimator.

Also for offers, I know there are a couple VN offers going around that some people are doing well with. One is to win a Galaxy S7 and another for an Apple watch. Keep an eye out for those and test them as well.

Also make sure you are using the ISPName and ISPID tokens in Popads so you can cut poor ISPs out. The ISP stats in Voluum are typically different than those in Popads.

Finally, if you can screenshot more stats that could really help as well. Carrier vs Wifi, OS and Device Types (are you testing tablets too?) are a good start.

Good luck on your journey!

---

It's good to split your campaign between iOS and Android. The reason behind this is you be bidding too much by targeting both and you could possibly be bidding lower if you separate the campaigns into two. This is something you'd have to split test, but based on my experience it's better to split between Android and iOS. Just a note on grabbing offers and adding them to your rotation, you'd want to make sure everything else is consistent. Offers are one of the most important aspects in the beginning phases of the campaign. You want to make sure that if you're adding offers you have to also test it with Angle 3 because Angle 3 might perform very well with your newly added offers.

---

Hello thien,

let me give you a few tips.

First of all, did you read this thread of mine : http://stmforum.com/forum/showthread...-of-my-methods
I think the strategy I outlined there should do well for you in this case and the advice Im gonna give you will be based largely on that.

So let's sum up what you have achieved so far. You know which offer works the best for you, and so far you didn't find a better one. You also have a few LPs that show similar results.
You didn't do much optimization when it comes to cutting placements, testing wifi VS carrier, android VS iphone...

So let's take what you know and build on it. I would suggest to start a few new campaigns now so you can take advantage of bidding separately for bigger segments.

The biggest differences are usually between wifi and carrier traffic, and Android VS iOS. This means, you will need 4 new campaigns at least. The truth is, bids can be pretty different from one carrier to another too, but let's not overdo it at this stage.

Use the same LPs and the same offer for all 4.

You have done some bid level testing too - I'm not sure what kinda bid you are running now with, but I see that the highest bid gave you the best CVR. The idea is to pick a bid that gives you the best CVR and best ROI, this needs some testing. Whatever bid you choose, set the same for all 4 campaigns for now.

You can also cut some placements right now, look for those that have horrible LP CTR - they will stand out from the rest at first sight. If you see a placement that ate a large part of the budget without giving you a single conversion you can cut that too - careful here tho, you tested some poor LPs and offers, so these numbers are to be taken with a grain of salt at this stage.

Look also at conversions by OS, brand etc ... whatever you can also pause at the source - sometimes there are smaller targeting segments that eat up budget and don't convert at all, so you might want to cut those too.

Once you got rid of the very worst segments, restart the campaigns and collect some data. It's the best to run non stop, some hours are better some are worse, so you need to get an idea about the performance across the day.

The next step will be to analyze your 4 main segments and play with the bid to maximize the performance of each. Then you can look at the placements again and cut the worst again.

Good luck and keep us posted

---

Hi matuloo thanks so much for your tips and suggestions. I am implementing now but I have some questions about this. So right now my daily testing budget is about $100/day, its actually less than that but I really can't afford anything more. So since I need to run these campaigns nonstop, does this mean I should increase my budget? Because if I split up the budget between the 4 camps, I don't think $25 per camp is going to last all day. Also how many days nonstop should I run these 4 camps? I am asking because these 4 camps have only been running for for about 6 hours now and my budget of $100 is already gone. I will load top another $100 but if this is gone in 6 hours I am not sure if I can continue the camp.

---

I suggested 4 to cover "everything" so to speak. If your budget doesn't allow this, you can choose just 3 or 2 or just 1. Then after a few days you can move to the other one. This is one way of handling this.

You can also set smaller budget per campaign - I'm not sure about popads now - do they allow even distribution so it's spread across the day or not? Many sources do.

You can also try lower bid so it lasts longer, try to change cap, pause the biggest placements for now ... there are many ways to tame the daily spend

As for how long you should be running these - for as long as you start seeing under-performing segments that you will be cutting. They need to start standing out - sometimes it takes a few hours, sometimes days.

---

UPDATE

Thanks again for the advice matuloo. Since the camps were already started, I didn't want to mess with any of the settings or bids because I read somewhere that can skew the stats because a higher bid can have placements that lower bids do not something like that, I am really not too sure.

I was not sure of the correct way to set up these 4 camps but right now the IOS and Android camps are both targeting wifi and carrier. The wifi and carr camp is targeting both IOS and Android. I have been testing with a .0025 bid this whole time so for these 4 camps I set it at .003 and ran it non stop for 48 hours.

I hope this was enough time to run it non stop because it spent about $560 and ROI dropped a little bit and that hurt lol. Here are the stats.


I looked into the reports of each campaign and I really do not see anything that I can cut. How much of my budget should be "eaten" before I determine to cut a certain brand, OS, placement, etc? Also what is a good CR for a lander on pops? As you can see the lander has an extremely high CTR but CR seems to never be better than 0.20%. I can't seem to get the camp better than 50% ROI. Any advice is appreciated.

---

Yes, this can happen sometimes, higher bid can bring you traffic from spots that would not send you traffic with lower bid - because the competition is too high there and people bid high.

So you have 4 campaigns now :

iOS - this targets iOS both on wifi and carrier connection
Android - this targets Android both on wifi and carrier connection

Then there are Wifi and Carrier campaign where both target iOS and Android. So what you just did is you are targeting the same traffic with the other 2 as you are with the first 2. You need to do it like this :

iOS Wifi
iOS Carrier
Android Wifi
Android Carrier

Why did you raise the bid for this actually? You could have continues with the original bid.[/QUOTE]

This was more than enough, you actually don't need to spend this much when you don't see it getting anywhere. It's normal to run at negative ROI but it must be heading in some direction. So for example you need to start seeing budget drainers in the data, or you have to see that some carrier is working better than the rest, or some OS is beating the other one. If all runs on the same level and nothing really stands out in a good or bad way, the problem will be somewhere else - offer, lander or traffic source.

The CTR of your landers is so high because you have some backbutton script on it which sends the users through the click url, am I right? You should send the backbutton traffic to a separate campaign because it is messing with your data a bit.

There are several things you can do now. This test didn't really help us, partially because you didn't set it right but we will still use it Try to analyze the data and figure out what combination gave you the best results - OS and connection type. Since these tests are pretty expensive, keep on using just this "winning" combination - we are doing this only to lower the costs, you can open it up anytime you see fit.

The next step would be to test more offers and probably more landers too, you need to find something with better performance. It's also possible you will need to switch GEO at some point, I'm not sure how well VN performs now, but sometimes there are GEOs that are no usable no matter what you try.

---

Wow this was one expensive mistake lol, at least I learned some lessons out of it. Thanks again matuloo. I will try to analyze the stats as best as I can but I may very well just switch geos. I guess I just really wanted this camp to work lol. If I were to switch geos and start over with a new camp, should I test wifi/carrier and android/os first, or should I test different bids first, or both at the same time? Meaning if I were to test 3 different bids I would have 12 camps set up instead of 4? Sorry if this question sounds confusing.

---

I know what you mean, 4 camps times 3 bid levels = a lot of campaigns. You don't have to do this. Pick one of the 4 and focus on that, you can open the targeting up anytime. So let's say choose Android/carrier or Android/wifi. Let's just say we are trying to limit the costs of the test by not targeting all 4 at once. Bid levels is something you have to test tho, but again, you don't have to start 3 campaigns at once. Test one bid level today, different one tommorow etc ... don't try to rush this, take your time and learn how to read data on the way.

---

Hey Thien,

Awesome follow along!

Looking forward to how this plays out.

I have a question though, Idk if this will help or hurt, but have you tested your LP speed?

From my personal experience Ive seen my worst landers that I was thinking about cutting turn GREEN overnight by simply compressing a few images and cleaning up the code a bit.

I usually test mine on Web page test

Just a thought if you didn't already optimize for speed.

---

Hey thanks davidito! Wow I feel honored that your first and only post since joining STM is advice for my campaign. To answer your question yes kind of, the only speed optimization that I do is compress the images through Kraken and run my landers on a CDN. I am not very technical so I have no idea how to "clean up" the code without "breaking" the lander somehow. Any suggestions on how I can achieve this? Thanks again for your advice.

---

Thanks again matuloo! I was under the impression to "split test" I should start the camps at the same time. I will definitely try this approach to limit cost. I do have another question regarding this. So say I choose one of these 4 camps to test, would I also be split testing multiple landers and offers at this point too in the same camp?

---

You are right, if you want to do a PROPER split test, it needs to run at the same time. But this can get expensive, that's why I'm suggesting to do it this way. You will basically just lower the volume and isolate one target right from the start, for example Android 3G traffic. The rest of the tests will run in the standard way, so multiple LPs and offers in this one campaign.

---

Caurmen wrote a lot of tutorials on landingpage creation, try to check some of them out, there might be info you need in them :

http://stmforum.com/forum/showthread...t-Ripping-Fast
http://stmforum.com/forum/showthread...-In-10-Minutes
http://stmforum.com/forum/showthread...With-Any-Phone
http://stmforum.com/forum/showthread...Actually-Works

And here is a good one regarding ripped landers from Vortex :

http://stmforum.com/forum/showthread...Ripped-Landers

And regarding speed, you might want to check these :

http://stmforum.com/forum/showthread...ter-Cheatsheet
http://stmforum.com/forum/showthread...s-Loading-Time
http://stmforum.com/forum/showthread...-Cost-Way-Ever
http://stmforum.com/forum/showthread...se-Your-Profit

There are more, so search for them

---

Home > Mobile > Follow-along Campaigns