Wakeboarder Trying to Get Some POP Working (31)
Hi guys!
So far I've tested 18 offers and only one is converting OK... Unfortunately this offer is for Cameroon and there is not enough traffic to start optimizing.
Campaigns are active for the past 4 days. Here are the stats:
Spend: $238.72
Revenue: $31.42
ROAS: -86%
Current my flow for testing offers is relative simple...
1) Ask AMs for offers: tier2/3; max payout $1.5; sweepstake; ... Win iPhone X or Samsung S9
2) Rip the the most popular lander from AdPlexity for particular GEO/offer
3) Spend $10-$15 per offer
I'm getting a lot of Android Webview traffic which is poor quality. If I exclude Android Webview in Propeller Ads dashboard I'm still receiving this type of traffic. Anyone else noticed this? Now I'm trying to detect these placements as soon as I launch the campaign and then I manually exclude them via zone ID.
Today I got 20-30 new offers from different proven aff networks.
Good news is that I can test offers much faster now because I already have landers in 10+ different languages inside
I hope some can help me with few questions, because I'm not sure anymore if I'm moving in the right direction. 
1) It's enough to spend $10-$15 per offer with payout below $1?
2) Most of the time I'm using two type of landers (Google Reward Program, Prizespin) in different languages. Should I test landers+offers at the same time at the beginning?
3) Right now I'm 100% focused on finding a good converting offer... It's this the right approach? I'm not optimizing TS, only blacklisting obvious bot placements.
4) I'm looking for an offer that will convert right away with min -60% ROI before I decide to put more budget for testing. Is that OK?
5) If I select CPM Cost model in
Your comments and suggestions will be appreciated.
Hey man,
Originally Posted by excavator
I have no idea how I misses this thread! Thanks for updating it to bump it to the top again!
I can't seem to see the picture you attached in the first post. In the future, if you want your image to appear nice and big, upload it to imgur.com and grab the bbcode link from there and paste it into here. Don't use the forum attachment upload cause it sucks (a limitation of vbulletin).
First of all kudos on testing so many offers! Below are some suggestions as well as answers to your questions:
-Try to rip and fix up more landers than just one. How good your winning lander is that you're using to mass-test offers, will directly impact on the performance of ALL those offers. So you need to test at least a few to maximize your chances of finding a really decent lander.
-Regarding whether 10-15 per offer is enough to test: It's a good rule of thumb for newbies, but every situation will be different, and basically you'd want to only test an offer for aa long as you can still see potential in turning a good profit after cutting stuff. Learning how to decide when to ditch an offer or campaign will require experience.
Also, it's always easier to test multiple offers at a time, so that you can cut inferior offers as they reach statistical significance.
-Although you CAN try to test landers and offers as the same time, I don't feel it would be the most accurate way to split-test. What I often find is that for many offers, the 1st conversion is easy to come by - what is called a "lottery conversion" where we'd have no idea when/if the 2nd conversion will happen. So if, when gathering conversions to help us test and cut landers, we use conversions from multiple offers, the results may be skewed.
This is why in the tutorial I suggest to run multiple landers and offers, wait until an offer reaches 2 conversions, then pause the other offers and just run that one offer to generate conversions for cutting landers.
After you cut down to a winning lander, you can always add the other offers back into rotation to split-test them.
-Regarding webview traffic, yeah that traffic doesn't convert. Your approach of cutting placements that contain lots of that traffic is the best way I know.
-Focusing on finding a good offer is 100% important! But make sure you have a decent lander first - not just use the one lander you grabbed from Adplexity.
Once you have an offer that looks like it's making enough of the total traffic profitable (some big placements, a major OS, etc.), continue to cut the unprofitable segments to reach profits.
Even while you're still testing offers and landers you can start cutting placements etc. that are draining the budget. You can always retest them after finding a good offer and lander.
-Your -60% ROI rule of thumb is great for newbies, and for the smaller of the tier 3/4 geos.
For competitive geos or big geos though, there may be much more room for optimizing from very negative ROIs to positive.
In most situations though, testing a good number of landers and eapecially offers, before going on a cutting spree for placements etc., would be a good idea.
Finding a good lander and offer to maximize profits from the traffic first, then cut the unprofitable parts. Don't take a mediocre offer and lander, spend a ton of money on cutting out big chunks of your traffic, just to end up making pennies from the leftover traffic.
Hope that helps! Looking forward to your next update!
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Amazing man, thank you! Works like a charm.
Originally Posted by excavator
Try to exclude android browser + android webview altogether, guaranteed they stop showing.. the weird thing is, i asked about this via chat + support ticket, but the answer was either the problem has been sent to tech department or try start a new campaign..
I found out this "browser's package" via random thought.. i tinkered with my phone, and i can only find browser app, so where does this android webview come from.. i can only presume that this webview maybe somehow related with android browser..
only one way to find out, excludes both android browser and webview, and there's the answer..
the only problem is this is the majority of traffic for a few geos, which means excludes them = risk throwing the campaign into a stale state..
i guess propellerads really sensitive if we optimize aggressively, or maybe it's just the way their algo works.. CMIIW
Originally Posted by vortex
Based on your feedback I've adjusted my testing flow a bit.
I increased the initial testing budget from $10 to $20 per camp and in the same time I will test 3 similar offers with the same GEO and 3-5 most popular landers for particular GEO.
Do you think it would be better to focus only on 1 GEO? That way I would build faster blacklist and exclude bot traffic.
Originally Posted by omelca
I calculated the % of Android Webview traffic per GEO for my campaigns. % is huge for some GEOs. See below.
Originally Posted by excavator
I calculated the % of Android Webview traffic per GEO for my campaigns. % is huge for some GEOs. See below.
I have pretty much the same data as u do, in general.. See that ridiculous large percentage just for this browser.. And it's just webview, the android browser must be somewhere near that number, but maybe slightly higher CVR, right? And the cost must be astronomical if compared to another browser, even with the one that converts much much better than these two..
is this one campaign or divided into each GEO? Idk if excluding both android browser and android webview right now will be a right decision.. the data won't be relevant anymore since they are all correlated.. <-- i'm not 100% sure about this too, will need someone with more experience to say for sure
TBH i'm always stuck in this phase too (with propellerads),
testing geo + offer -> optimize placements -> faced with android browser and webview, again.
Exclude both browser if there's enough traffic left = suddenly the other browser dive down in performance, when it seems promising before excluding. but sometimes this is not the case, for example in Indonesia, exclude them both, traffic still flowing, conversions too.. but the payout is very low :'(
Different approach, go straight to optimize/cut zoneId = daily budget gone with the wind - or should i say the android browser duo, low CVR, HIGH cost, scratched my head, sleep.
Maybe amy can shed some light after reading this, i wish..
Numbers above are for different campaigns. Every geo represent one or max two campaigns.
When you're testing offers... are you matching pre-landers potential prize and the prize from original offer?
Take a look at this pre-lander. User have an option to win iPhone X because other 2 prizes are gone In this case will you drive to an offer where the prize is iPhone X, right?
Regarding focusing on 1 geo: when first starting out, that MAY be a good idea, because then you wouldn't be wasting money learning the same lessons from multiple campaigns.
However, once you've optimized a couple/few campaigns and know what you're doing, launching in multiple geos can be an efficient way of finding something promising faster. I would highly recommend this approach. If you're just running one camp at a time, you'd be holding your breath for that one campaign to succeed, which is not a good thing at all (you'd be more proned to being driven by emotions when making campaign decisions).
As for prize on lander matching prize in offer: Yes of course it needs to be the same prize. Put yourself in your visitors' shoes and it should be obvious.
Trying to make the transition from lander to offer page as seamless as possible can do wonders to CR.
As for the webview targeting: Let me ask around to see if anyone has a better solution. Stay tuned!
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Thanks, Amy!
Regarding webview targeting... I'm excluding Android Webview and Android Browser and it looks much better now.
Originally Posted by vortex
Btw Amy, can you suggest some material reading regarding in depth optimizing vertically or horizontally?
Thanks
Originally Posted by excavator
Do you have any campaign which getting most conversions from these 2 browsers excavator? I wonder what's the deal with these browsers, maybe some publishers bombing propellerads with sophisticated bot using these particular browsers?
wish you good luck!!! noticed your post since im also wakeboarding for years.. started also kiteboarding two years ago.. and yes, affiliate is the way to live free. been there done that for 10+ years.. but casino industry.. wanted to try also smth else... in case you will ever need investor and if your project would run faster because of some help with €, lmk. =)
good luck!!
Originally Posted by omelca
Yes, it's working for now. No, I got a few conversion from these 2 browsers, but the CR/CVR is terrible.
I think this type of traffic is coming from spy-tool crawlers (adplexity, anstrex, ...) and bot placement.
Originally Posted by quepao
Thank you for you kind words. Kiteboarding is also in my goals list but first I want to be really good at wakeboarding.
Luckily the money is not the problem at the moment, but I would be more happy to get some insights/comments from someone who is in this industry for 10 years. Will send you a PM. Thank man!
Here is my workspace that keeps me motivated.
CAMPAIGN UPDATE
Over the weekend I was optimizing the system (
I believe I've optimized the system for testing and in the next few days I will just test offers & landers.
I post an update with more details in a day or two.
I've asked my lovely contact at PropellerAds about this - will reply here when I get an answer.
(I'm also wondering whether it'd be possible to somehow make use of this traffic. Webview cannot execute javascript. So perhaps by targeting JUST webview traffic at a low bid and testing landing pages that are designed to convert WITHOUT using javascript would be a worthy exercise?)
We post update with more details in a day or two.
https://stmforum.com/forum/showthrea...considerations
https://stmforum.com/forum/showthrea...ow-to-optimize
Other stuff I could think of: minimizing server speed and lander-loading speed, and using automation tools (custom-made by yourself, or commercially-available ones such as theoptimizer.io).
Do you have anything specific in mind when you say optimizing vertically/horizontally? (It may be good to start a separate thread for that discussion - but please update in this thread and link to the new thread if you do start one.
)Amy
Originally Posted by vortex
I like the idea about testing this type of traffic using LPs w/o JS. I will skip this for now because I don't want to lose my focus.
Initial results of the new testing system looks much better today.
Let's see what happens until tomorrow.
had to say, love your setup dude! still i cant imagine working without mouse for larger period of time =) (i spend a lot of time in adobe CC programs..)
keep the spirit alive and wish you good luck on your journey!! btw, where is your wish destination to live? anyway, if you ever come to EUROPE lmk we can visit Hipnotics or some other cool place! =)
peace!
Originally Posted by quepao
Thanks man!
The setup with wakeboard was more for the fun and picture. But it's behind me in the office all the time. Actually, I think I'm already in my wish-destination. Super-small, beautiful and safe country... Slovenia, Ljubljana.
I know Hipnotics... it's one of the most amazing wakeboarding places. Can't wait to visit this park, really. Yeah, let visit some cool places. Will DM you.
CAMPAIGN UPDATE
After adjusting the system for testing landers and offers campaign performance increased a lot. I've soon found a combo (offer+lander) that looks promising.
ROI is -41% after 44 conversions and I'm excited AF.
At this point now I'm not testing so many offers anymore, but looking for a ways how to increase ROI. Here is my plan how to optimize to achieve break-even.
1) Payout bump
I've asked my AM for a payout bump. I told here I'm close to break-even and slightly higher payout would help me send more money to the offer. Waiting for her reply.
2) Other networks, same offer/GEO
I'm looking for the same offer in other networks where I'm in good relationship with AMs. I would like to compare performance and see if I can get even higher payout.
3) Placement blacklisting
So far I've blacklisted only 1 placement. I don't want to be too aggressive with blacklisting because of the volume.
4) Other traffic sources
A few days ago I already opened an account on all major pop networks. I will make a deposit, setup
5) CTR optimization
CTR on lander A is much higher (x2) than on B... but CTR is 17% vs. 68%. I'm not sure why's that. Will do some tests and try to increase the CTR to get more click from this lander with good CR.
6) Placement killer calc
I will start using placement killer calculator that Amy mentioned in her top-notch guide.
Yesterday I tested two different bids in PropellerAds. I noticed that after I start new campaign with higher bid ($2.2) spend on the old one (bid $1.5) dropped significantly. Is this normal?
I would appreciate any comments on my scaling plan.
Here is my fav track that I'm listening when I'm in the wakepark... It's the same feeling when I refresh a
Hooked on a filing
EDIT: At which point would you recommend me to open a whitelist campaign? I have a handful of placements with 3-6 conversions... should I just leave everything and slowly blacklist non-performing placements? What do you suggest guys? Thanks!
QUICK UPDATE
Yesterday through the day campaign was performing at around -35% ROI. I've installed SSL on Amazon/Cloudflare setup and today in the morning I noticed a significant drop in performance. After few hours I figure it out that problem is because my
Language
When I was creating a campaign flow I create a rule which redirects traffic with Language = English to English lander and Language = Malay to Malaysian lander.
As you can see from the screenshot below, Malay is performing much better than English.
I'm redirecting now all traffic to Malayan lander.
Is this usual that local language performs so much better?
Here campaign performance for the past 3 days.
Tomorrow I will decide if I will kill the campaign or leave it running. In total, I've spent over $300 on this campaign... which I think it a bit too much.
Do you think that it's worth to spend more on this campaign that is around -50% overall?
Great! You're doing a lot of the right things. Here are my suggestions:
1)Look through your placement stats to see how many placements are in profit with at least 2 conversions, and calculate how much profits they are making per day on average.
If that number is smaller than day $5, then I wouldn't even bother to optimize further.
Also look at other traffic segments like OS and browser. If there is a big segment that is doing very crappy ROI, cutting that may increase the ROI for placements overall. Otherwise, what I said above applies.
You need to make sure there is enough profit before cutting out the unprofitable parts.
2)If however, you do feel you that you have enough profitable placements, then just cut unprofitable placements. If you feel you have a promising lander and offer already, then don't worry about maintaining traffic volume.
3)If on the other hand, you don't feel that your current lander and offer combo is making enough od the traffic profitable, then you should split test more landers or offers, or both (test one first, find a new winner, then test the other).
How many landers and offers have you tested so far for this camp?
4)I would not suggest scaling to other traffic sources before you've tested enough landers and offers. Otherwise you may just end up losing money on multiple sources.
The exception is when you're running in a geo with a ton of traffic from a ton of placements. In some cases, you'll need to cut a ton of placements on every traffic source you scale to in order to see profits.
But in any case, it would be necessary to split test landers and offers until you find a promising combo, before scaling.
The better your combo is, the more traffic you can make profitable, and the less money you'll need to spend on cutting placements before you see green.
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Forgot to mention: That kill-whitelist calculator is overkill for pop camps. That calculator will help you to cut with quite a high level of accuracy, but the flip side is it will require a lot of data before it can reach a verdict on whether to cut or whitelist. With there being so many placements in pop camps you'll go bankrupt cutting placements according to that spreadsheet.
I would suggest to cut placements at 2x payout in loss. You can even cut at 1x payment in loss or less if you're dealing with a huge geo that has a ton of traffic and placements, and losing a lot of money, so need to limit losses fast.
Once you find a good offer lander combo, you can always retest soms of the placements that are previously blacklisted.
Regarding whitelisting: For pop camps, blacklisting usually works better. Whitelisting would often result in you getting miniscule traffic volume, but you're encouraged to try.
Regarding staggered bid camps and how higher bid camps can take traffic from lower bid camps, yes that's a common phenomenon. In that case, the better bid-testing approach would be to start with a low bid, cut placements, increase the bid, cut another batch of placements, etc., until the daily profits decline, then revert to the previous bid.
Amy
Sent from my SM-G930W8 using STM Forums mobile app
Originally Posted by vortex
Originally Posted by vortex
Originally Posted by vortex
Originally Posted by vortex
Originally Posted by vortex
Important takeaways from this camp
- don't get too excited too fast.. be patient
- if it make sense... test local language and english
- always triple-check when changing technical stuff (tracking, hosting, etc.)
- test more offers, test more landers
Next steps
Simple... test more offer and more landers.
I will still update this follow along until I don't make a great campaign. Your comments and suggestions are really valuable and incredibly useful. I'm happy to be part of this community and I hope someone else will find this FA useful. Thanks guys!
Special emphasis on testing offers.
As you've mentioned, there are usually just a few different popular landers floating around, whereas there are usually a lot more offers.
The trick is to spend as little as possible to find something promising, and THEN focus on optimizing and scaling the promising stuff. I know it's easier said than done, especially when you don't yet have the experience to have the confidence required to tell the difference. But your odds are improving all the time.
A lot of what you're finding out is very valuable as well - testing languages for example. It pays to look up which languages are spoken in which country (either that, or just run some traffic, then drill down to "language" in your tracker to see which languages are most widely-spoken in your target geo), then set up tracker rules to direct the right language to the right landers.
Tricks like these may not give you very massive advantages individually, but once you have a few of them, they'll collectively give you a major advantage over affiliates that DON'T bother.
And to make money, you don't need to be better than the best guy - you just need to be better than some of the competition.
Amy
Yes, over the weekend I'm planning to test at least 10 news offers and 5 landers / every offer.
Yesterday the campaign I was working on over the week turned out in some small profit. It's not scalable because I'm targeting only local language... but it's more valuable in terms of motivation and data about placements. When I found 4 small offers like this one... I will be close to $100 profit / day which is amazing.
I found out that bottleneck in my testing flow are AMs because they're giving a few offers at a time. I'm trying to get a spreadsheet of all offers that are relevant for me (sweepstakes, electronic devices). How do you guys handle this? Do you ask AM every day for new offers?
CAMPAIGN UPDATE
Old Malaysia campaign is now doing approximately $30 /day in profit without touching it. I've copied the campaign to PopCash with exactly the same settings and ROI is 50%. Now I will just leave the camp on both TS and just check once per day if it still making a profit.
New campaigns
Before the weekend I managed to get approved numerous offers from my AMs.
Here are the overall results for the past 3 days, including profitable MY camp.
I found one campaign from CL that's converting great but the overall daily cap is only 100 conversions for the whole network. Because the payout is low it doesn't make sense to put too much effort into it. I sent a message to AM if he can do something about it.
I tested additional 5 offers without good initial results and I killed them after $15.
I've noticed on several affiliate networks that travel sweepstakes offers are popular. I'm thinking to test some of them. Right now I'm doing only sweeps for iPhone/Samsung and I'm limited and slow with testing offers.
What do you think about this? Should I stick only to one sub-category and put my focus only to landers/geos/placements that are related to electronic devices or is a good idea to test other categories as well?
It's really interesting how challenging is to find a winning combo. Here is my list of variables sorted by importance:
- offer that convert
- good lander
- offer cap
- scalable GEO
Next steps
Again, test more offers and more landers.
Looks good man, well done!
Thanks @diplomat! 🤙
Sent from my iPhone using STM Forums mobile app
Hi, Rick!
I’m glad to hear that! 🤙
No, I’m not running pops anymore. One of the main reasons why I moved away was because campigns with high ROI didnn’t last long. Carrier billing type of offers became too regulated and with SOI offers I never found some big success.
But it’s true that I did not put too much effort into SOI offers. Afaik the main source for SOI offers is FB and to run sweeps offers there you need to be ready to handle account bans. For me personaly, this is part of the game I didn’t want to enter.
Pops are definitely a great entry point to learn the ropes and get used to tracking, split-testing, optimizing traffic source, LPs and so on.
Hope this helps!
Cheers,
waleboarder
Sent from my iPhone using STM Forums mobile app
@Wakeboarder I would suggest two manual things which works for me. Try having more Pop Network like Pop Ads, Adcash instead using only propeller ad and pull out the performing zoneid from Adplexity for specific offer. With this approach the conversion will come faster. If you can spend more time try and go deep into publisher and see what kind of campaigns they are running that will give you an idea of publisher's traffic quality.
For Lander I suggest try using a common man image with prize pic. People link with these things quickly. Hopefully this should work for you.
Thanks
Bharat
Home > POP / PPV / Redirect > Follow-along Campaigns