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

---

Amy's 40 Day Guide (Voluum) (31)

---

Hey guys, this is my follow-along the Amy's 40 Day Guide using Voluum. My only experience before this is 1$ guide, I also kind of figured out basic affiliate glossary but some things are still somewhat confusing.


I want it to make journal-alike where each day i'll note things I did, what went good, what went bad, what could be improved/done better. My goal is to get first campaign to the green during the course and to learn more about whole affiliate flow. I think it will be cool to recall this post later on, to see all of the progress, and to have laugh on mistakes that are undoubtedly going to happen.

Okay, so without further ado, let's get started

<this thread was inspired by follow along made by @GregMorrison 's https://stmforum.com/forum/showthrea...g-(with-Binom) while back ago>

DAY 1 - 2020.09.04

Things I did

I mostly focused on applying to different affiliate networks and traffic sources.
Created account on Mgid, AdCombo, ZeroPark, PropellerAds, ClickDealer, Advidi, Gotzha, Visto
Already had account on Mobipium and PopAds.
I also purchased new domain that will be used for tracking on namescheap

What went good

1. I had about an hour-long skype talk with Voluum salesperson that presented me whole platform workflow, I was thinking about Redtrack before but this definitly helped change my mind.
2. Got approved on most of the networks

What went bad

1. Purchased Adplexity Mobile - I don't really know if I'll use it in next few days, before the day 17
2. I tried to register on Mobidea but got some weird "Proxy Connection" error, contacted support regarding that issue, still waiting for response
3. Topped up Zeropark account using PayPal with different email that the one that I used to register Zeropark account which ended up in short account suspension (had to confirm both emials and use CC instead)

What could be improved/done better

1. Paperwork and taxes - I want to store invoices at the time of receiving it on GDrive, kinda postponed it to the next day and had to look for all transactions that were made.

Questions after Day 1

1. How are people utilizing Adplexity?
2. Would I cut cost after setting up self-hosted tracker? Would this be worth the effort?
3. Non-english GEO - is it better to use native language or to use English? What is the best place for hiring native speakers to do affiliate translations?

---

Great in the beginning to swipe the landers but I could not justify paying over $100 on a monthly basis when the angles of the landers does not change a lot. At least for the sweepstakes. I have been barely using it.

Good luck with this journey! Wish you a profitable campaign!

---

DAY 2-4 - 2020.09.05 - 2020.09.07

Things I did

I've set 4 campaigns on Mobipium and tried launching another one of Smartlink PN Landings while waiting to be approved and reviewed by Clickdealer.
The traffic source used was Zeropark, but it turned to burn down money really quickly.
The smartlink offer literally burned 20$+ from the time I noticed email notification till the time i logged in into the platform to see how it rolls. Overall profit from 30$~ spent on traffic was around 1.5$. Seems that it won't break even without LP.
Then, I went to STM and read somewhere that it's actually better to use Zeropark while scaling campaign and not at the beginning, so want to verify that using other traffic sources.

Went through Amy's Day 9, 10, 11, wanted to start day 12 but prioritized writing this post and leaving day 12 for tomorrow.


What went good

1. Got some conversions, already have around 6$ at Mobipium account
2. Decided to set separate Google Chrome context dedicated to affiliate marketing to separate everything from personal stuff. Already had separate email address set up since the day 1.
3. 1Password subscription turned to be very effective in the vast majority of the websites I had to sign in during the last week. I recommend using password manager to anyone just starting as it saves time and keeps everything more secure in case of potential data leak from one of the services.
4. Today (07.09) got a chat with Clickdealer AM and was finally accepted after longest interview ever
5. Created account at Mobidea, couldn't do so before because of some strange Proxy Connection error but changing wifi network fixed the issue
6. Launched two 10$ campaigns for one offer from Clickdealer and one from Mobidea, got only 3 conversions from both but Voluum is starting to make sense at last.

What went bad

1. I don't know why but i initially thought that affiliate networks and traffic sources are working 24/7, or at least during the weekend 8-5, this assumption pretty quickly was broken when I couldn't get anything accepted/verified during the weekend


What could be improved/done better

1. Setting up campaigns on Mon-Fri, spending Saturday/Sunday solely on reading STM Forums and educating myself
2. Taking more action on forums.


Questions after Day 2-4

1. Where to host landing pages? AWS?
2. For anyone bit more technical - will it be worth to set up git repository with landers and some kind of auto-deploy chain to save clicks and time? Something like setting up lander locally, adding it to repository and pushing to see it listed on the web
3. Somehow related to previous one - is it better to use subdomains for landers or to keep each lander on separate domain? Does this make any diffrence? Former would be easier to set up for auto deployment with apache virtual document root

---

@plutus I can already tell that this follow-along will be great - actually it already is!

Good insights. And congrats on getting accepted into all these networks - especially Clickdealer which isn't exactly easy to get into.

Will respond to some of your questions and comments below.

1)Regarding Adplexity: This tutorial would be a great start: https://mobile.adplexity.com/tutorial

There are many ways to use this tool. Here are the main two that come to mind:

a)Find lander styles that work. If you sort by "Received most traffic" in a recent timeframe, and rip the most popular landers, you can't go far wrong because you know that many people are getting conversions with them or else there wouldn't be so many people using them.

Of course it would be great to innovate based on them, but when you're first breaking into a vertical and/or a new geo you need to use popular landers to at least establish a baseline. After that you can always split-test more lander version of the best lander, and even test completely new lander designs/styles.

b)Find out what's hot right now. For example, filter by the country/countries of your choice (a geo that's been recommended by an AM or a traffic network rep, or just pick random geos that are tier 3/4), sort by "received most traffic", analyze say the top 20 or 50 or 100 results to see which verticals/offers are doing the most volume and on which traffic network and even which affiliate networks the offers are from. Then you can go chat with AMs from those networks to see if you could also test them.

Note that just because other affiliates can make those offers profitable doesn't mean you can too (they may have already optimized their campaigns and bumped to the highest payouts), but at least you'll know which verticals and offers are doing well in that geo at the moment which can really narrow down the number of things you need to test.

The downside is that you'd be following the crowd, because by the time an offer "makes it on the radar" in spy tools, it's likely at least half way to saturation. So you may not want to limit yourself to testing those exact offers, but instead sign up to more affiliate networks that have offers in the same geo+vertical and test those as well.


2)If you wish to focus on pop and when you start to run bigger volumes, yes that can rack up a big bill on trackers that charge based on events, at which point it could make sense to switch to a self-hosted solution. But right now you're not quite at that stage yet, so I'd suggest to just keep using Voluum for now as you're just starting to get comfortable with it. You can always cross that bridge when you get there.


3)Of course it would be better to target each geo using the most prominent language used there. For starters, ripping landers that have received a lot of traffic should suffice. Or, if your lander doesn't have a lot of text, you can keep the sentences very short and to the point, and use google translate. Of course google translate won't give you the best quality, but for initial testing it's often good enough.

If you really want to give offers the best chances of success, but not wanting to spend too much money, check out fiverr. Pick translators that have lots of reviews and a near-perfect star-rating and you can't go far wrong.

Also: Some languages are spoken in many geos, so if you pick one of those languages, you'll be able to test many geos with a single set of landers. For example (in addition to English of course): Spanish, French, Arabic, and Portuguese - see here for a list: https://en.wikipedia.org/wiki/List_o...icial_language


It can be good to do all this research upfront to narrow down to a few geos and verticals you want to test (in spy tools and by chatting with AMs and maybe other affiliates), rip and prepare a bunch of landers, sign up to affiliate networks that have the right offers, then test away on a good-quality traffic source. Optimize camps that have potential to profitability, scale ones that show promise to other traffic sources. Rinse and repeat. That's it in a nutshell.


Many traffic sources are actually staffed 24/7, some with longer campaign approval times during weekends. Affiliate networks not so much, but some very dedicated AMs can be reached during the weekends as well, especially if you're running volume.


1)There's a step by step in later lessons of the tutorial, using AWS yes. But of course you can also get your own hosting. It's just not necessary though because you'll be using a CDN anyways so it doesn't really matter much (or at all) where the files are hosted - they'll be served by the CDN.

2)I'm the least technical person to be answering this one, so I won't. It would be the perfect question for @jeremie though if he sees this! Please consider starting a new thread with this question to get more feedback.

3)If you're not running higher volumes yet, having one domain at a time for all landers should suffice. You can have a generic domain name with niche-specific subdomain names, such as win-free-iphone.bestdamnoffersonline.com. The concern is that google can ban the domain, so if you're running high volume traffic, it may be good to have separate lander domains to mitigate risks (i.e. when one domain goes down you don't end up with 0 conversions for the day).

Hope that helps!



Amy

---

For me, it is worth the time. I would say it depends on your skills, and that is not the first step I would take. I am implementing this setup these days:
https://stmforum.com/forum/showthrea...-lps-using-aws

Be aware that AWS S3 + CloudFront is not recommended if you do pops, for the same reason Amy mentioned for the cloud tracker. It is charged by data transfered and numbers of HTTP requests, which can be costly when doing volume.

---

@vortex Thanks, bookmarked Fiverr and gotta follow along with Voluum to the point I get volume that is out of the events range, funny tho that Voluum sounds like volume and you have to switch to something else when getting VOLUME

Besides that, I'd like to thank you for this tutorial (and upcoming 2020/2021 one!!!) I really appreciate amount of work you had to put in to get it up and running, and also the dedication you put into newbies like myself. This is really helpful.

Thank you for linking that thread, that was exactly what I was looking for, will save it for later.

DAY 5 - 2020.09.08

Things I did

Started reading explanation of PropellerAds on Day 12 but did not finish it yet, decided to take some more action and set up another campaigns, went for PIN submit and AV offers (first one on ClickDealer, second on Mobidea). I've bought pack of ripped landers from Adplexity for 1$ and set up my first landing page on custom domain for AV offer, did not go CDN way yet. There was also one campaign from Day 4 that turned out to 9 conversions at the end, I tried optimizing it by adjusting browsers and switching device to android (8 out of 9 was from android) and cutting of different languages by whitelisting Azerbaijani, Turkish, Russian. Set both campaigns to 10$ and went to bed.

Woke up this morning at 5AM, was dreaming about affiliate marketing and green columns and couldn't resist to check Voluum app to see how it rolls, and... well. There was no conversion, and only few clicks with -2.5$ cost, got some additional data tho.

One thing is weird in PropellerAds - when editing campaign I noticed that this auto-optimization switch was enabled and it ended up setting CPM bid in ZA to over 5$ which exhausted most of the 10$ in no time, still need to figure out how to set bidding up

What went good

1. Launched another two campaigns
2. Started ripping off landers from Adplexity, and boy, this thing is addictive! Want to do it on daily basis as I don't plan to get another month subscription
3. Got first LP up and running, it was verified by my ClickDealer AM and I sent some traffic there, Voluum is cool with this /click link on landers
2. Optimized my first campaign, but...

What went bad

1. ...optimization did not bring new conversions, instead I ended up receiving none
2. My new LP received close to none clicks - around 10, I don't know if this is because of misconfiguration or something else


What could be improved/done better

1. I want to set up 1 campaign daily while following along the tutorial to keep consistent data flow, feel like I spent too much time selecting offers and did not spend enough time to go further in the tutorial
2. Had more questions in mind before but I forgot them - gotta note questions as they arise, to avoid forgetting them


Questions after Day 5

1. How many conversions are enough to optimize campaign? I feel like this 9 mentioned above is close to none, would need like 10x more to use it as valuable indicator
2. Is it worth optimizing campaign that got 10~ conversions for 10$ (0.12$/conversion) or is it better to go on with different offer?

---

LOL it IS quite ironic! And thank you for your kind comment - I will channel my gratitude into making the tutorial better.


A few things:

-Definitely figure out CDN before continuing. For pop, page load time is important.

-Congrats on getting conversions!

-Adjusting browsers - what's the justification? Feel free to show screenshots of stats. (Instructions can be found in @twinaxe's post: https://stmforum.com/forum/showthrea...-In-Your-Posts)

-Switching device to just android because 8 out of 9 were from android - but how is the traffic distribution across ios/android? If android comprises 8/9th of the total traffic (or close), then you can't justify cutting out ios yet, for example.

-Cutting different languages - again, what's the justification? (Again please feel free to show stats.)


Yup next time please turn off auto-optimization and try manual bidding.

Also: ZA traffic is expensive. I would advise to start with geos with cheaper traffic. ZA is way up there with tier 1 geos like US/CA/UK in terms of CPMs.

And the lack of traffic is probably due to the additional restrictions you've put on the campaign. Did I read correctly that you've whitelisted Azerbaijani, Turkish, and Russian for a ZA campaign? If so then there's your problem!


Good stuff! If you do your research (as I've outlined in the previous post) to decide which verticals and geos to focus on in the next few months, you can just rip a shit-ton of landers for them in a day and not need to use that tool for a while. Landers don't stop working that quickly.


Both due to lack of traffic as discussed earlier in this post.


It's a great habit to be launching every day, but when you're still learning, it can be better to launch fewer campaigns - you need to learn how to throw and catch a single ball before learning to juggle.

That way, you don't repeat the same mistakes across multiple campaigns simultaneously.

Once you're comfortable with testing and optimization, you can come up with plans to mass-test offers by launching lots of campaigns. That day will come. But trying to run while still learning to walk can cost you money and leave you frustrated.


You can't really look at a single factor and make a decision. This is way too general a question. I would encourage you to skip ahead to the lessons on optimization - they're overly complex and I'll make sure to improve on that in the updated version, but they should give you a much better idea on what to look for when making optimization decisions.


Excellent work so far! You're taking a LOT of action which is priceless for a newbie, but now may be a good time to slow down a bit and focus on reading more before launching more campaigns (or relaunching the same campaign since it was on auto-optimize).



Amy

---

DAY 6-7 - 2020.09.09 - 2020-09.10


Things I did


Okay, so I decided to follow your advice @vortex and read much more about AM in general and spent whole day yesterday purely on tutorial and linked threads. I quickly noticed, when going more and more deeply into the topic, that I kind of wasted 20-30$ for campaigns that could not be successful without a proper lander. So well, here I am at the second day, decided to go on with non-aggresive sweepstakes, picked up 2 offers in Thailand and even found two nice landers already. I am currently doing web-dev for living in my 9-5 so I skipped Day 14-15 (only part of getting skilled with html, js, php) cut out some crappy js out of landers, replaced tracking urls and went straight to setting up AWS and S3. Following @jeremie advice I went to different path than CloudFront - decided to test it with CloudFlare which is free and as far as I know it does not charge users per view. Setting everything up turned out to be much easier that I expected, followed this cool tutorial: https://miketabor.com/how-to-host-a-...nd-Cloudflare/


There is also SSL included so everything runs on https YAY!


So at this moment I have 2 landers ready to go, still need to figure out some kind of auto-deployment to S3 to make everything setup faster (not a priority). I decided to go with single domain and subdirectories with landers - need to update directory structure as with current naming convention it would be super easy to guess other landers path and I would rather not want to get mine stolen landers to be stolen (inception huh, nice movie btw). Overall I feel like am starting to getting somewhere and not just blindly going ahead. Last two days were fun


What went good


1. Gained additional bulk of general AM knowledge
2. Decided to go with sweepstakes when following along this tutorial
3. Found and cleaned two promising landers, still need to find additional 4 for each offer
4. Did set up custom domain for landers with S3 + Cloudflare
5. Voluum is really nice - this feature with linking /click page instead of the offer link on lander opens so much possibilities and variations that campaigns can be set up without pain in the ass
6. Stopped wasting money on campaigns that are not converting because of missing landing page
7. Just by doing this follow up posts I noticed how my writing skill is improving, this is really nice


What went bad


1. Maybe because I'm newbie but I asked AM on multiple networks about specific offers and received bulk, looking like copy-paste of offers mostly not relevent to the thing I requested, to be honest I don't know if this is just how the way is or I need to earn more money to get some (not so) special treatment.


What could be improved/done better


At this stage nothing comes to my mind as I've mostly did theoretical part of tutorial, this section will be HUGE soon enough when I start testing landers and optimize it


Questions after Day 6-7


1. bulletin is kinda retarded, I try to copy previous post format and it's bolding everything up without possibility to unbold it. Maybe you guys know if there is a way to use Markdown or something similar instead?

---

Just so that no one is misled, I said that CloudFront is not recommended for pop. I do recommend it for other types of traffic.
My personal setup is:
- S3 / CloudFront for html pages (because I need lambda)
- S3 / CloudFront / CloudFlare (in front of CloudFront) for images and JS. I have the two together because CloudFront control the cache expiration and the setup is easier than S3 to CloudFlare.

If you connect S3 directly to CloudFlare, it is interesting to note that by default, S3 does not assign an expiration date to uploaded files (unless your deployment setup indicates it to S3) and therefore S3 will not send cache control headers when accessing a file. As by default, CloudFlare respects the cache-control headers, you could end up with files that are not cache that long. And if you have for example 10 landers with 10 files each and a 6-hour cache decided by CloudFlare, that will be 12000 hits to S3 in a month.

To avoid this, you need to setup a "Page Rule" for all files, indicating to CloudFlare to force the Edge Cache TTL settings to 14 days or one month. (Not Browser Cacher TTL).

---

@jeremie thanks, just set up Edge Cache TTL to month with Cache Level set to Cache everything, added additional rule to add https rewrites as I don't want to serve my landers from http (lame). I will see if this work in the next few days.
Current page rules:

---

Which Voluum plan do you have? The Discover one does not support SSL on custom domains, only on the dedicated domain.

---

Actually, Discover one, one of the Voluum sales representatives contacted me and offered annual subscription for half of the price, for 39$/month instead of 69$ + got SSL Domain as a gift. I decided to go on with that, and since Zeropark events are not counted in the 1M events counter I'll probably use Voluum just for that later on if my volume increases, besides that, I'm not hosting landers on my tracker domain Does http/https rule also apply to landers? I've already added one lander with https to Voluum before, this is weird

---

I am just talking about the custom tracking link that you set up in Voluum, which according to what i read from some here, can not be https in the Discover plan.

Therefore, if you define your landers with Https and the custom tracking url that users click on is http, it is going to be a problem.

---

ah, okay, but both domains are with SSL so this shouldn't be a problem.

DAY 8 - 2020.09.11

Things I did

I mostly focused on fixing up LPs that I ripped yesterday, focusing on the two offers that I found in Thailand - Iphone 11 and Samsung S20. Managed to fix five LPs for Iphone11, three are in thai, two in english, got one lander for samsung s20, and one that is kind of universal - simple swipe to unlock that redirects user to the offer. I also set up automatic deployment to S3 from local git repository using DeployHQ that turned out to be free to use if managing single project. Besides that - I'm working on directory structure that will help me manage dozen of landers efficiently, so far ended up with the one attached below.

My reasoning is following:
done - landers that are fixed and ready to go, but not deployed (some sort of landers backup)
fix-in-progress/current - lander that I'm currently adjusting/fixing
pending - raw, ripped landers without any adjustments done besides unzipping to separate directory after downloading it from Adplexity
public_html - this is a directory that I'm pushing to S3, decided to go with following structure to make everything clean - VERTICAL//OFFER/LOCALE/NUMBER-HASH, added random hash to offer to secure myself from anyone crawling my lander's domain and downloading everything without much effort.
rejected - LPs that did not convert well but are fixed and some adjustments might make them work and roll

I also wrote simple PHP script that recursively iterates over public_html directory and generate all of the offer urls that are deployed to S3:


This is still stone age, and I need more data to improve this process even further. I'm planning to set up one global javascript file for all LPs with all of the AM scripts, like getting query args or redirecting to offer after pressing back button. Next step would be internationalization and replacing raw html files with templates and adding translation files as .mo or .csv. Deployment would then put the same lander with all the different languages in separate locales directory (mentioned above). I don't know yet how freelance translators work but I think that .mo or .csv would make their work much easier, need to clarify that before for sure. But the question is - will this be worth it? I noticed that the same LPs are used with multiple languages, so probably yes, will see in the near future.

I ended up on Day 26 of Amy's tutorial and started to optimize load time of my landers, but I did not progress far ahead in the post, leaving it for tomorrow.

What went good

1. Created git repository with landing pages and set up automatic deployment after pushing changes.
2. Fixed 6 LPs, replaced malicious js and tracking URLs
3. Figured out initial directory structure for working with landers efficiently

What went bad

Since I mostly focused on coding I did not encounter any major difficulties today

What could be improved/done better

1. LPs deployment as stated above but this can be done later on, the way I set everything up today should last for couple of weeks

Questions after Day 8

1. What about copywriting? This is a skill that I should probably train, do you have any book recommendations? One that I saw Ca$hvertising on other subforum
2. Does the same LP in multiple languages can convert?
3. What is the workflow with freelancers if it comes to translations? Do they get zip with files or list of the sentences to translate?

---

The SSL certificate must be issued through Voluum. If you have put a SSL from Cloudflare, it won't pass the CNAME redirect. That's not obvious but i got the problem recently while uploading files to S3. Be aware.

---

DAY 9 - 2020.09.12

Things I did

Today, I launched first campaigns with multiple landers, I'll post some stats below and explain what I did.

Offers are Sweepstakes - Iphone 11

Stats from today:


First offer that I tested with 5 LPs - 3 in Thai, 2 in English

I quickly decided to cut off English as both landers received 10x less clicks than Thai ones

something weird happened with Voluum, I had more clicks than visits (Thai - Iphone 11 002)

After budget ran out I decided to create another campaign with two LPs that converted the most (001, 002) and cut out 003

and I found a winner 001, and 002 yet again had more clicks than visits, I probably failed to clear all malicious scripts, i don't know what could be other reason for that to happen

I then decided to go on with LP 001, and ended up with 1 conversion after spending 4.5$, I also completely messed up bidding price as I tried to change it too quickly, after launching this campaign I noticed that PropellerAds starts sending traffic after about like 15-20mins from getting approved, and every time I've set something up there I instantly panicked if there was no traffic at all after 3-5 minutes and incremented CPM for 0.2-0.5$

Stats of all three campaigns:


Then, out of nowhere PropellerAds chatbox popped up with list of the countries that are converting well and with a question if I want to know which offers also do convert well so I asked. Got bunch of links to the offers from Zeydoo and one of them was and Iphone one. I decided to give it a shot, offer stated that Egypt is converting the most.

Offer lander looked promising so I set up campaign both on Voluum and PropellerAds and decided to test both English LPs. When setting everything up I noticed another Iphone offer which could be tested so went on with 2 offers and 2 Lander.s

Stats of second campaign:


and steps I've taken:
1. I quickly decided to cut off second offer as I received 2 conversions from first, and 1 conversion from second.
2. Soon enough decided to cut off first lander, it was 2:1 in conversions.
3. All that was left was Lander 002 and Offer 001, I excluded zones that was not so cost efficient and set campaign budget to 20$, was checking stats to see if I'm getting any conversions with the new set up but it ended up received only one. I decided to stop the campaign when cost reached value close to 10x revenue.

What went good

1. I saw huge improvement in conversion rate over setting up campaigns with direct links.

What went bad

1. Wasted around 10$ on incorrect campaign setup on PropellerAds, set CPM way too big
2. I feel like I gave up too quickly on the campaign #1 and wasted too much money on campaign #2

What could be improved / done better

1. Getting this daily posts written one hour earlier, I'm half-asleep every time I'm at the moment of writing it.
2. Selecting more offers to test instead of going on with single one
3. More patience when starting new campaign - I won't adjust bidding price so far above the average no more
4. fixing LPs - one of the incorrectly adjusted lander prolly screwed up my campaign stats

Questions after Day 9

1. Was in that particular case better to go on with campaign #1 instead of giving it up for campaign #2 which turned out bad?

---

Sweet!

Yes this will definitely help with the "testing multiple geos with the same set of landers" idea we talked about a few posts ago.


1. Cashvertising is good. Also: We have a whole subforum dedicated to copywriting, that you can check out:

https://stmforum.com/forum/forumdisp...amp-Psychology

Also: Study landers in spy tools! If you see a particular lander style a lot, figure out WHAT makes them work so well. Same goes for landers that have received a lot of traffic. For pop landers, you need to a)REALLY capture people's attention in a fraction of a second - they're bascially interrupted from browsing whatever it is they're browsing on the website when your ad appears, so the page needs to have something that will make them pause to read the rest instead of clicking close right away; and b)make the text VERY short and to the point - make visitors want to follow through with the intended action, and explain why they have to do it NOW, and make any instructions stupid simple.

2. Of course! As I've mentioned in a post above on how to do research in spy tools, just look for landers that have received the most traffic, and do this for multiple geos, and it should be apparent which styles are consistent winners across geos. If you're planning on testing multiple geos with the same set of landers, this would be the way to go.

3. Each freelancer has their own preferences, but most will follow your instructions as long as they're not too inconvenient for them.



Amy

---

Amazing! You got some conversions - very nice start!


Not surprising of course. I would advise that as a newbie, try to always use the local language on landers. Yes I HAVE used English landers to quickly test lots of countries to see which ones have potential, which does save time but burns money - which is why 99% of the time I would use the local language.


Looks like there's double-accounting going on somewhere...


No need to start a new campaign. Next time, just go into Voluum campaign settings and remove that lander from rotation.

That way you retain all the campaign data, rather than splitting the data across 2 campaigns.

And did you use a stats calculator to check for statistical significance before cutting?


Definitely looks like it's a problem specific to that one lander.


It's OK - all the "messing up" is the only way you'll gain experience. Trying different things is the only way to learn.


1. Eek! You didn't check for statistical significance first did you? Another few minutes and offer2 could have made another conversion to even up the score. Another few minutes more and offer2 could have converted better than offer1. This is why we need statistics - to take out the guesswork. Of course nothing is 100% accurate, but picking one offer over another just because it made one conversion more is definitely not the way to go.

2. Same as above.

3. What criterion did you use when cutting zones?


It's actually hard to tell whether you've spent too much or too little.

Whether you should continue to run a campaign or not, depends on a)how much room there is for improving your ROI, and b)whether that "room" is enough to turn the campaign green.

Example: Say you spent $10 testing a $1 offer, and it converted 3 times (-70% ROI). Do you continue?

Let's say you were testing 2 landers, and one of them appeared to convert twice as well as the other. Theoretically if you only run the best lander, you'd be looking at increasing the ROI from -70% to -60%.

And let's say you drill down into placement (zone on propeller) stats, and see that a big placement has spent $4 without converting ($2 to each lander). So theoretically if you cut it, you'd be looking at increasing the ROI to -33%. And when you get the ROI to -50% ROI or above, you stand a good chance of getting to green by running more traffic and cutting more placements.

And that was just an example to illustrate what I said above regarding whether there's enough room for improvement to get green. There are other optimization opportunities you can look at, for example: cutting wifi / carrier traffic, cutting certain mobile carriers, cutting IOS/Android, etc.

And often, you'll just have to burn money to run enough traffic to most of the bigger placements to see which ones you need to cut and which ones to keep. But once you've done this, the next offer or lander you test on the same targeting (e.g. geo) on the same traffic source, you will have already cut the worst placements so you'll have a head start.

And if you find an offer or lander that converts better than your previous campaign, you could retest some of the previously-cut placements, except for the really hopeless ones that have burnt a hole in your pocket in the past - those you can put in a permanent blacklist.

Campaign optimization isn't as easy as this of course. Every case is different and it's hard to provide rules of thumb - which is where experience comes in. But hopefully I can provide you with a good starting point.

Now let's look at an opposite example: Say we're looking at the same $10 spent testing the $1 offer which converted 3 times. But you're already using a proven lander which was the winner in a 5-most-popular-landers-you-found-on-Adplexity split-test, and when you drill down to zones the spend was pretty much scattered across all the placements (i.e. no placement zapping the budget and not converting), and you were targeting a single carrier in a single geo and only android phones because that's what the offer accepts. In this case, I would think twice about continuing to throw money at this campaign.

I would test more offers and landers first, to make sure I have a good combination, before taking another stab at that geo again. Sometimes you just have to burn money to find out the good and bad placements - as mentioned above - when running in a geo for the first time on a traffic source.

Or, I would just test another geo altogether. This is why it would be wise to test several geos at a time if you have the funds - so you can find the most promising to continue testing and optimizing.

And remember that testing needs to come before optimization. Don't just take one offer and one lander and try to optimize your way to green - most of the time it won't work and even if it does you could do it for cheaper. Test landers and offers to find a winning combo first, then focus on optimizing. And between testing offers and landers, focus on testing more OFFERS - way more.


Please feel free to drill down to various stats and post them! As mentioned above, every situation is different. We need to see how much room for improvement there is before making a decision.

And if you didn't cut landers and offers using statistics - then you'd need to re-run the testing I'm afraid. It's a great thing though that you're running offers with such low payouts - this is allowing you to collect quite a bit of actionable data with minimal spend.

Good work! And please don't feel obligated to write a detailed report daily (as much as we love to see details and are very happy to answer any questions!) Only post for as long as the writing helps you - e.g. to organize your thoughts, and help answer questions you have. If you know what you're doing on a given day and just want to be left alone so you can have more time to test stuff before posting more stats and asking more questions, that can be a good thing as well.



Amy

---

Okay, I had no time not to update this follow along as I was totally absorbed with the landing page development setup I am preparing for myself for the future, and also some nice things happened - more info below

DAY 13-14 - 2020.09.13 - 2020.09.14

Focused 100% on the development setup and ended up with following:
1. decided to go with twig templates -had some background with this templating system before and it turned out to support pure javascript rendering in twig.js
2. after some struggling I managed to set up gulp to compile twig templates, and compile scss + minify css output, it basically looks something like this:


json files are the translations, {{ test }} tag from index.twig is replaced with the string from json, neat thing about this is that I can store any html markup (with <script> tags included) in that json and it is parsed and internationalized and it also stores my Voluum tracker url, this will make adjusting it later on when my domain gets blacklisted MUCH easier.

I can add as much languages for the same template I want and everything is compiled and copied to dist folder with locale suffix, so for the screenshot example it ends up as index-en.html and index-th.html and both htmls got template tags replaced. Okay, but lets cut off tech stuff for now, in the next day I tried something new, want to know why? Sign up here for only 4.99$

DAY 15 - 2020.09.15

My AM from ClickDealer (really nice guy btw, if u see this someday Tomi - I'm glad that we are working together) saw that I applied to multiple Thailand Iphone offers and recommended me one that I decided to try with previously ripped and prepared landers.

I added one lander that I translated using Google and set up campaign on Voluum & PropellerAds. After spending about 10$~ I got about 7-8 conversions, cut out different offers that i side-tested and turned out to be conversion-less and cut out 2 out of 4 landers that I was testing.

After another 5-6$ I noticed that one of the landers is receiving fairly more conversions so I cut off the second one (@vortex - just saw your comment here about A/B testing, will definitely try this out next time) and continued campaign with 1 lander and 1 offer.

I noticed that some of the zones are not converting at all and receive fairly amount of views, so decided to cut them. I was cutting out next zones until I hit 20$ campaign budget limit, had about 11 conversions at this time (0.25$/conv). This was the most expensive campaign that I had but decided to give it a shot and added another 20$, from this point on things went pretty quickly -

I started receiving more and more conversions from the same zones and none from the others and cut out basically everything that was not converting.

I ended up with about 10~ zones not blacklisted, I set up max bid on all of them, for the ones that converted the most (2-3) I set fixed price to 7.5$~ , for the rest I set it up to 4.5$. I initially thought that this would burst out all of the remaining 20$ quickly but it was running for quite some time, I wasn't sure what else can be adjusted so was just waiting to see if this setup works. Gotta post some results below to finish up this wall of text and entertain u a bit, here we go

Campaign final result:


zones:

browsers:

browser versions:

Great, I'm happy that I actually did it before reading your post, getting all this data somehow changed my point of view, got a pretty decent starting point for the Thai sweeps offers.

This is not the end of the day yet - I got more offers recommendations and one was from Poland for Media Markt voucher - I quickly ripped and fixed one voucher lander and set it up, went to Propeller and set up another campaign, but...

Well it turns out that maybe this lander wasn't polished enough but I ended up receiving 1% CTR comparing to 40% CTR in the previous offer from Thai, somehow got 1 conversion tho, decided to cut campaign after 3.5$~ and add some extra stuff to the lander to make it more appealing, gotta try to add some alert and additional js to make everyone stay longer.

Polish Lander:



Thai Lander from the first offer - this one and some slight variations are literally everywhere on Aplexity, want to adjust it somehow to be different because I think this is the factor that can make me break closer to the green



Polish campaign results:


Noted, I'll try different geos with the Iphone offers

DAY 16 - 2020.09.16

Just started getting things done in AM with this post, I'll apply to some ClickDealer offers from different GEOs and that's basically it for today, I'll have more free time in next days so there will be ACTION!

Questions after Day 13-15
1. wtf is wrong with that PL CTR? Gotta double check if I didn't left any malicious redirect script in the lander
2. given the stats - is it actually worth it to go on with the Thai campaign further? I'm kind of running out of ideas what can improve it besides setting up new landers to test
3. what is the best way to choose GEOs to test? I selected Thailand based on PropellerAds traffic volume in that country but just wonder if traffic amount should be the main indicator

---

DAY 17 - 2020.09.17

I found a nice working LP for IPhone 11 on Adplexity and decided to convert it to work with my tool. I ended up spending whole afternoon on clearing it and making it multilingual, replaced all texts with template tags, I also found cool script in obfuscated javascript that did malicious redirect if brand_type = 'Galaxy', since it looked to nice I decided to adjust it to my Mobipium Smartlink, this will be really funny if I somehow manage to get any free traffic there in case if someone rips it without applying much of the work to fix it.

For anyone reading this without much of coding experience - if you see something like this in your lander better be aware - atob, btoa, eval, new Function() are one the keywords that should light a red lamp in your head
ff59 = "&u=" + ff57 + "&v=" + ff58;
for (var idx1 = 0; idx1 < ff37.length; idx1++) {
for (var idx2 = 0; idx2 < ff37.length; idx2++) {
try {
var ff38 = atob(ff35(ff37[idx1], ff37[idx2]));
let ff41 = new Function(ff38);
ff39 = idx1 - idx2;
ff40 = idx1 + idx2;
ff41();
} catch (err) {
}
}
}

I ended clearing everything up and deploying and decided to give another shot to Thailand campaign after copying some texts and google-translating rest. Final result:





DAY 18 - 2020.09.18

TESTING!! I launched the campaign with previous LP and the new one and here are the combined results from the end of previous day (launched it after midnight) and from today
Since I had some funds deposited on Zeropark before I decided to give a shot to the same campaign with different traffic, I think that I'll save rest of the funds when I find a campaign that ends up profitable on PropellerAds to scale it at Zeropark, cuz it turned out really nasty.





And overall result since I started this follow along:



I am still not sure if this offer is sufficient enough, I want to adjust my spinning wheel LP and add survey before displaying potential Iphone rewards, but am not completely sure if this won't be an overkill. I also started reading more about sweepstakes in general, found a @vortex vertical guide and ended up at part 6 and also found some old threads from like 2015ish with more of like general knowledge that can still apply to present day. Both landers ended up almost the same I want to do further adjustments to this campaign by doing following:

1. Adding 3rd lander to the collection with survey before rewards are displayed
2. Cutting of tablets as it did not convert at all -

3. cutting out Huawei as only got only one conversion for about 1400+ clicks



3. whitelisting Chrome Mobile and Facebook App as rest of the browsers are not converting at all



4. cutting out AIS and DTAC as both are not converting so well




I will post the results as soon as I finish adjusting new LP. So far it's going pretty nice, drilling down all the stats and seeing all possible optimization paths gets me really excited. Till the next time


p


EDIT: YAY!! STM LEVEL 2 !

---

DAY 19 - 2020.09.19

Yet again I focused on improving landers and tool and ended up creating common js shared across all of the landers with scripts like dates handling, timer script, onpopstate, onbeforeunload and malicious obfuscated redirect in case someone would steal it. I also added support for Google Translate and setting up the same LP to new GEO is a matter of one command right now:









I ended preparing the landers at 3 a.m today - decided to set up 2 campaigns with bunch of new thailand iPhone 11 and 12 offers and went to bed. I also did all the optimizations mentioned at the end of previous post before starting them.

DAY 20 - 2020.09.20

I woke up this morning and to my biggest surprise I noticed that my campaigns are actually almost green, this is what I saw:



I quickly applied optimizations based on fresh data, here is what I changed:

1. Previous screenshot was of 3 offers that I tested, so I A/B tested all:



...and disabled both B and C offers

2. Next, I checked landers stats:



and also did A/B




and decided to go on with lander A

I left the campaign running for the whole day, and since it was doing fairly well for most of the time at -25~-30% ROI I decided to test it on ZeroPark and PopAds, I just launched campaigns on both traffic networks and there is still much to optimize, here are the results as for now:



Questions:
1. I ended up limiting everything to the level that I received 5k visits for campaign that was running for 12h~, should I remove some of the websites from blacklist? I'll try another lander with survey but just wonder what else can be done so this can somehow be scaled in case if it's turn out successful - that's one of the reasons I went to other traffic networks as rest of PropellerAds zones wasn't performing at all.

---

DAY 21 - 2020.09.21

Added budget to Propeller and launched another campaign with the same offer, I also spent few hours preparing new lander, I feel that I'm wasting way too much time polishing each LP instead of quickly fixing it and testing if all that work is even worth it. So all day went by, here are the stats:





DAY 22 - 2020.09.22

I created new wheel LP with confetti pop after two spins and tick sound while spinning, created account at Clickadu and ended up setting up this thailand offer again. I also applied to multiple different multi-geo push subscription offers and launched 4-5 campaigns on propeller, and so far Costa Rica looks promising. I just realised that previous screenshot of total spent was wrong as it had incorrect filter applied, excluding archived campaigns. Real stats as for now:




Most of this revenue is on ClickDealer (45$+) and since my AM is really cooperative I decided to stay there until I reach payout revenue (500$) to avoid splitting it on multiple networks. So far it looks terrible, I'm leaving this thai campaign for night and if it won't get around -30%~ ROI I'm going to cut it off, previous days mislead me, this was so close, highest ROI i saw was like -20%~. My AM contacted me today and even increased PO from 0.34 to 0.38 but I don't really know where to optimize it. Gotta go back to theory as I mostly focused on action and wasn't reading that much.

I just started Singapore campaign as all of the offers have less than 0.20$ PO and I'm totally fucking everything up till now so I guess even lower payout will give me more space to do so


Questions:
1. Is adding a sound to the LP valid improvement or should it better be left out without it? I kind of thought that this would be cool adjustment but I don't know if its not too much intrusive

---

DAY 23 - 2020.09.23

My biggest fuckup day so feel free to laugh. I launched the same campaign with some minor Propeller Ads adjustments and after like 20$~ spent I realized that I didn't even check the flow of the new offer and it turned out that my lander have completely different texts in todo-list after spinning is done, my offer is SOI, email submit and there was an info about filling up shipping address. I really feel so stupid right now, have been running this lander for close to week and wasted another like 80$ on this campaign in span of last 2 days and 20$~ on the push subscription offers.

Stats from today:



Stats from last 2 days:




and overall:



I'm stopping TH camp, prolonged it WAAAY too much

Things I learned (really hard way):
- double checking flow of each of the offers launched and adjusting every lander to make it fit correctly - lander with wrong offer description is not a good way to go
- setting bids lower - this fucking TH, i had it running mostly on Propeller and ended up having 7$ CPM zones - setting bids way to high for campaign with not verified lander, I also went to GEO that seems to be super competitive, second most listed country in Adplexity, after limiting up zones and other stuff I ended up with suggested CPM of over 40$ (LOL)





- setting the same campaign on multiple traffic sources without getting it to green on the first one - this is stupid, I just spread my cost even further because of that

I'm not giving up, gotta just do things in less chaotic. Plan for now before launching another campaign:
1. Creating some kind of campaign workflow, when to cut what so I won't end with campaign that cost me over 100$ ever again - this is just classic example of being scared of sunk costs. I kept this TH running way too long
2. Reading up Follow-Alongs and re-reading Amy's tutorial
3. Improving copywriting and changing landers - I think that my masturbation about getting perfectly coded lander was so fucking stupid, wasted so much time and effort for setup that improved whole flow just a little bit, adjusting one lander for whole day is not a way to go
4. Choosing new GEO, still staying on sweepstakes but I'm thinking about changing vertical as I understand correctly so far to get conversions you have to be super aggressive and that's not a way I want to work at all, I don't want to force people to convert, would rather be persuasive enough to convince them to do so, not sure if there is much of a difference between the two

---

Hello! Did you send your campaign details to our Support Team? They can help you find the solution to this problem.

---

no, I did not contact Support Team regarding that, will definitely do it next time

DAY 24 - 26: 2020.09.24 - 2020.09.26

I did yet another attempt to TH, went full on with Facebook Browser only and it turned out lander that I pointed to thousands of people before to was inferior to the other one tested in this campaign. I think that some translation tweaks could break this campaign even but decided to try new GEO - Indonesia.

I set up fresh RON campaign on Propeller with previously used landers translated using Google Translate and wasn't expecting much without native translation but ended up with 21 conversions, still -90% ROI

Some stats from the last 3 days -

Campaigns:



TH landers:



IN landers:




and overall stats, BLOODY RED:



I'm not touching my Zeropark account until I find any campaign profitable this could've been 100$ less in cost.

DAY 27: 2020-09-27

I'm mostly focusing on reading at the moment, my paperback Ca$hvertising landed on Friday. My plan for next few days is to read Ca$hvertising and Tested advertising methods, do Fiverr translation to TH and IN of landers that received most conversions, will test new ones in the meantime - I see giftbox variant a lot on Adplexity and haven't actually do any testing with that.

Until next time,
plutus

---

DAY 01-30 2020-09-01 - 2020-09-30 First Month in AM

I'm going to post summary from my first month doing AM. I learned a lot. I'm still improving my copy. I'm still testing offers with bloody red results. Green? You wish.

Since the beginning of September I launched 40 campaigns for the ClickDealer offers with payout varying from 0.10$ to 0.38$.

I mostly used Propeller Ads as my traffic source. Also did some test run with Clickadu. AND wasted about 100$ using Zeropark in two-days span (this one needs a fast-blocking hands, will remember that)

Overall stats from September:



and the campaigns stats:



So MUCH more to learn!

DAY 31-33 2020-10-01 - 2020-10-03 Poland

I found nice looking PL offers on ClickDealer for the local grocery store voucher.

I quickly decided to create landing page from the scratch for them.

Things that inspired me the most was the surveys LPs. I also mixed some elements of spinning wheel variant of Iphone sweepstakes.

I wanted it to look as close to the advertised vocher's store as possible. Final look of the first lander:



After first launch I got it to -30% ROI - this is by far the most I got from the initial start without any prior optimization done.

I decided to improve it even further by adding facebook-like comments and better-looking alerts using sweetalert js library.

Final result after improvements:



I'm still waiting to see if the job I did was right or wrong. I did cut the most non-converting zones (3-5x payout). I also prepared different variants of the lander with slight changes in copy and layout - 4 LPs total.

---

Wow lots of progress has been made!!

Sorry I didn't get around to posting - will strive to post more often!

All the lander coding stuff sounds intriguing - wished I had more technical know-how to appreciate your setup more! Sounds badass though - especially the ability to switch to a new lander domain at will!

One question: When cutting zones, did you cut based on overall stats across all offers and landers, or based on the best offer+lander combination?

Zones that appear to be in loss overall can turn out to be profitable for the best offer+lander combination.

Of course to save money you can cut zones based on overall stats in the beginning, but once you've pinned down the best offer+lander, drill down into offer > lander > zone stats to see which zones you need to send more traffic to in order to tell whether it can be profitable for the best offer+lander combo.

Also: I would start a separate campaign and just target the FB browser, with all but the very worst zones blacklisted.


I commend you on taking all this action!!

However - as you've already learned - throwing up a single lander, which you haven't run in the past and don't know if it's even good, is no way to test an offer adequately.

When you get bad results with this approach, you wouldn't know whether it's the lander or offer that sucked.

Whereas if you were testing 5 popular, well-optimized landers, you can practically guarantee that there's at least one good lander in the mix - so if you don't get conversions, you can pretty much point your finger at the offer.


Yup that lander CTR definitely looks low! But see that's my point - if you had several different landers running, you'd have a WAY better idea where the CTR should be - as much as the CTR can vary widely from lander to lander, you'd still see what the average/typical range should be. This lander would have stuck out like a sore thumb.

Yes the Thai campaign is worth dissecting more! Targeting just the Facebook browser would be worth a shot for example. You can also drill down into other stats such as wifi/carrier. Also note what I said about evaluating overall stats vs. stats for your best offer+lander combo - the latter is way more important!

As for geo selection - if your AM assures you that an offer is doing well, and it's in a tier 3/4 geo with a low payout, then it's worth testing out!

You can even narrow down to a specific niche in a specific geo by 1)asking all your AMs for offer recommendations, 2)checking spy tools to see which verticals are doing the most volume in each geo, and choosing one or more geo+niche combos by combining intel from both of these sources.

Or even: Do research on Adplexity by looking through tier 3/4 geos that speak the same language - e.g. English/Spanish/Arabic - then pick a type of offer that seems to be getting significant traffic volume - e.g. iphone11 sweeps - then go to all your aff networks to ask sweeps for those geos. That way you can test more offers using the same 5-10 landers.


I've tried that in the past actually - including code that siphoned a portion of the traffic to my campaigns if they were ripped. I soon stopped doing that due to personal reasons, but the few landers I did that for continue to bring in a bit of traffic for a long time afterwards.

Mega thanks for the tips regarding sneaky code! I know you must be very busy - but if you ever find the time to write a post on how to identify and remove sneaky code, that would help out a lot of members!


The first one - worth trying for sure!

The second - you can't really draw that conclusion, since very little traffic came from tablets.


Good stuff!


In-line with my suggestion before - I would even target Facebook solely as it looks the most promising and most likely to reach green quickly.


Fair enough!


I love your reasoning and your approach, and the fact that you're taking a lot of action! One warning: Cutting so many things at once can be like throwing the baby out with the bathwater! If you were using Funnelflux as the tracker you could drill down to unlimited levels of data to figure out exactly what ROI will "theoretically" result by cutting all of these traffic segments at once, but even then it would take a lot of math. A simpler way would be to cut the most-negative-ROI segment first (that consists of a big enough chunk of the total traffic to be worth cutting), then running more traffic to see where you're at before cutting further.

For example, after cutting one major segment, you may be able to optimize to green by cutting a few more zones (i.e. smaller segments). That way you don't "overcut" to result in a small stream of traffic you can't do anything with.


\

Haha congratulations! The first couple of levels are easier to progress through - then it gets harder and harder.


Amazing!! Next you can try some of my suggestions to monetize more of the traffic.


Good going with your use of statistics to arrive at that conclusion! I would have waited until the "probability to be best" to reached at LEAST 90% though, preferably 95%. But to each his own!


Yup answered above.


Please consider my suggestion of preparing one set of landers for the same offer type in multiple geos that speak the same language - can save a lot of time this way.


Good call on sticking with ClickDealer if cash flow is an issue - they have a wide selection of offers.

You're being too hard on yourself - taking massive action and FAILING FAST is the only way to move forward quickly. But yes, alternating between doing and reading may help!


That's something you can test. I don't remember now why I stopped using sound on my landers - in the recent couple years they seemed to have disappeared from ripped landers. It's been too long - maybe it was because the performance between sound vs. no sound was moot, or maybe it was because traffic sources were rejecting them. I don't even know if it's easy to make automatic sound work universally on devices anymore. If you find out please enlighten us!


Again - way too hard on yourself.

Try to set aside a disposable test budget for each month, and write it off in your mind without expecting to make back one cent. I'm not saying to do this forever, but at least for the first few months while you're learning.

Then, aim to spend this budget to learn as much as possible - and this learning WILL include "fuck ups"!

I'll always remember this one night when I first arrived in a new country where I would stay for the next few months (the affiliate lifestyle). I had just gotten off the plane and was exhausted. But I felt "inspired" to set up a campaign for an international offer where I was targeting many geos. Then I went to bed right after.

When I got up in the middle of the night, $2k+ was spent with no conversions. Turns out I messed up the tracking - either I wasn't using the right aff link or didn't set up postback correctly or whatever. Had to restart the testing and write off the $2k+ spend as a loss.

So....yeah, you didn't do so bad!


Always a good idea!

Which goes back to the importance of NOT "overcutting". And I wouldn't bid so high UNLESS and UNTIL I've weeded out the "bad" zones at lower bids first.


Exactly!


1. Can consider my suggestions above and retry.
2. May be able to pick up a few more tips!
3. If the improvements will help you save time on all future landers, then it's worth it. But I know what you mean - keeping in mind the 80/20 rule.
4. Unfortunately, pop and push traffic almost necessitates being super-aggressive. If that's not your style, I would suggest to run FB or adwords or native. I stopped running pop for that exact reason.


Why? You've spent so much money on TH already - any new offers and landers you test against your current best offer+lander can be done for cheap! And you can do all this testing on only the best zones (it will take longer yes, but you can start other campaigns in the meantime)!

New offers and geos and verticals are ALWAYS tempting, because it's the unknown and you always have that hope in the back of your mind that it can convert so much better than what you're currently running. Although this can turn out to be true, don't forget that when you jump around a lot, you'd be perpetually stuck at the data-buying stage and skipping the making-your-money-back stage. Focus on conducting tests in order to learn the nuances of each traffic source setting, and to refine your testing and optimization approach. These are the things that not-too-many affiliates will bother doing, and by focusing on learning instead of profits, you'll have a leg-up over other affiliates in the long run - because they'd be depending on luck whereas you'd be relying on your knowledge of how the traffic network behaves, and your tried and tested methodology.


This is very wise!


Sounds like a plan!


I would celebrate this instead of focusing on the red. It would practically take a miracle for a newbie to end up green in the first month.

You've learned a TON with those few hundred dollars!


And now we wait!




Amy

---

At the beginning I was cutting it based on overall stats but then switched to cut only on the most recent information - I'm still figuring out workflow, I'm currently testing if it's good to create one test RON campaign without really messing with settings (besides blacklisting total crap like -3xPO) and go further from that place with the new campaign with winning offer, blacklists and additional adjustments based on the tracker data.

Unfortunately my Voluum subscription does not allow me to drill down in more than ONE stat in report and at this stage I'm really far from upgrading it to higher tier for 100$/month more just to get that - would rather spend it on traffic sources.

I actually did native TH translation of spinning wheel lander but it did not convert as much comparing to the landers ripped from Adplexity. I think I didn't go aggressive enough tho. I tested aggressive style in PL campaigns and ended up on -30% after first launch (my previous post)


I tried it before but as far as I got was -20% ROI and it started to fall down until budget was out.

Yeah, I think I'll go back to the stage I with the knowledge I got because I'm just drifting cluelessly around.

My Adplexity subscription has ended but I ripped like 100+ sweeps landers and didn't even unzip most of them - I'll start with that again.

Do you have a Tier 3 GEO's list? I used the one on Propeller - https://propellerads.com/blog/what-i...ld-you-choose/

and I checked every single country from the tier 3 GEO column and found like 2-3 offers on ClickDealer.

This really confused me because everyone says everywhere about shitload of tier 3 geo offers and I barely found any.

Maybe ClickDealer is not a good place for such an offers?

Maybe should I test Multi-geo offers with Tier 3 traffic? I'd really appreciate little advice here

This is sad to read, I hope that you at least had a good rest that night

Noted. Like I mention above I started to mess less while testing to get additional data

Sure! Will try my best - saw your post yesterday and started taking notes already.

Yeah! I'm happy that I managed to start taking action. I'll wrap this follow-along in few days (4 days to day 40th) and start separate follow-along on sweeps + pops.

DAY 34-36: 2020-10-04 - 2020-10-06 Poland II


After lander improvements PL campaign went green for like 30 minutes that made me SUPER excited:



But ended up at like this:





I found out zones that look promising tho:




Plan for the next few days is:
1. Cleaning up more landers ripped from Adplexity
2. Making PL lander even more aggressive and split-testing it alongside some most popular sweeps landers
3. Testing more offers
4. Finishing up Ca$hvertising - 3/4 done - I'm taking shitload of notes as this book contains so much valuable knowledge that can be applied to the AM industry

---

It may be better to not start a separate campaign, because then all your zone stats etc. would be scattered across multiple campaigns, and you'll need to combine the stats manually to figure out what zones to cut etc. But it's up to you.

Developing a consistent workflow isn't easy - you will get there as you run more campaigns.


I get that.

In that case you'll need to spend more money on testing. For example:
-Throw all offers and landers into a campaign, wait until one of the offers makes 2-3 conversions.
-Pause all offers but this one, and pause the campaign. Restart the campaign on a new day to use this offer to test landers.
-Use statistics to cut landers down to a winner, but only take into account data from the new day onwards (because before that date, stats were from ALL offers).
-When a winning lander is found, pause the campaign and restart the next day, using that lander to retest all offers. (Again, when using stats to cut offers, only look at stats from the next day onwards - this is why it's important to keep a campaign journal so you'd know what you did when.)

Alternatively, use Binom - at their low pricing they over-deliver massively. Or use funnelflux - free month + 50% off for the first half a year.


This doesn't surprise me anymore. Sometimes google translate beats out translations from OHT. You just never know until you test.


Did you include back most of the zones when you did this? When targeting too narrow, the traffic and quality can dry up. And it's also common for campaign performance to go downhill over time. Sometimes just cloning the campaign on the traffic source to restart it will do the trick. But I leave that to your discretion.


Yeah, I think I'll go back to the stage I with the knowledge I got because I'm just drifting cluelessly around.

My Adplexity subscription has ended but I ripped like 100+ sweeps landers and didn't even unzip most of them - I'll start with that again.


I wouldn't go by any list - it's not important which countries are considered to be what tier. The general observation is that DEVELOPING countries tend to be less competitive and have cheaper traffic than DEVELOPED countries. Do a search in google and you'll find lists like this one: https://isge2018.isgesociety.com/reg...ing-countries/

Multi-geo offers can be fun to play with, especially international offers, because inevitably there will be countries that are under-saturated that you can monetize, i.e. the less-popular ones or smaller ones.

Clickdealer has a ton of offers - I'm relatively sure they have offers in tier3/4 geos. You can also check out offer aggregators like Offervault.com to see which networks have a lot of tier 3/4 offers and contact them.


Sounds great!


Did the original campaign go downhill? Or did you split them up into several smaller campaigns?

If it's the first case, look at hourly stats to see if you have enough data to start dayparting.

If it's the second case: Often when you target narrow, the traffic source algo will assign less importance to your campaign and stop giving you the volume or quality traffic. Can test both though.


Sounds like a plan!



Amy

---

Yeah! I'm aware that it will take some time.



Haven't tested it like that and it looks like it would let me get as much out of my current Voluum plan as is possible. Thanks!

I've already paid for 1 year of Voluum. I plan to switch to the higher tier when I get to the level of x-xx$/day campaigns. Binom is the one I'll give a shot when my Voluum subscription ends.

This surprised me A LOT! I was expecting it to get some higher CTR/CR but it turned out to work the opposite way.

I used STM tool to register on multiple networks and I got approved on Gasmobi. To my biggest surprise I found the same offers that I applied for PL in ClickDealer are listed in Gasmobi with higher payout (0.80EUR compared to 0.55EUR). Is it possible that someone listed offers on ClickDealer just to get conversions from the offers listed in different network? Something like middleman?


Original campaign went downhill. I was split-testing 4 landers variations and cut of 2 that got no conversions. Overall campaign received few conversions at the start and later it decreased. Today I launched it again with the winning lander and new, more aggressive variation and it ended up with no conversions after 18$~ spent (stats below) in previously converting daytime. I'm switching to the something else as it looks like that PL offer burned out. Got bunch of zones to blacklist for PL so there is a gain.

DAY 37: 2020-10-07

I launched PL campaign today on and it received no conversions at all. I also did another attempt on TH with previously working lander, just for the fb browser without success.

Stats from today:



I'm fixing ripped landers, one of them is 1 billion google search award. no idea if that kind of landers will work with sweeps in general. gotta test it

---

DAY 38-39: 2020-10-08 - 2020-10-09


Some more PL & TH:







DAY 40: 2020-10-10 - Summary


So... 40 days have passed.


Time has come to wrap this follow-along and move on.


I learned A LOT.


From Day 1 newbie struggling to even set up campaigns to Day 40 newbie successfully setting up 40+ campaigns.


I haven't reach my initial goal of getting green campaign (had one for 30min~ but I think that does not count).


But I think I reached one of Amy's goal for this tutorial - I got much more confident in setting campaigns, testing offers and landing pages.


I'm still figuring out workflow that will suit me best.


I'm 100% sure that I won't be doing pops & sweeps for entire life - this vertical and traffic type combo is just not my style.


Overall stats from Day 1 to Day 40:





Question: Where to go from that place?
Answer: More pops + sweeps (for now)


I'm starting separate pops + sweeps follow along and I'm planning to stay there until I reach consistent x-xx/day campaigns to cover all monthly AM costs like tracker, STM etc.


I plan to follow recently published @twinaxe article on getting x.xxx/month doing penny campaigns - https://stmforum.com/forum/showthread.php?49127


Next step will be getting 2k$+ profit from sweeps over next couple months and trying Facebook advertisement.


That's it for now. Thanks for everyone that helped me during this journey.


You guys are

---

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