40 Days of learning (43)

Hello everyone,

I started today my 40 day challange.

I went through all the basics:

• SignUp with Mobidea
• Signup with PropellerAds and adding my first funds
• Setting up my custom tracking domain
• For the tracker I thought I start with Mobidea's integrated tracker to keep steps quick and easy
• I choose my first 2 offer from Nigeria (Video and Game)
• I started my first campaign that is just in this moment pulling in the first impressions

Yes, PropellerAds spent the first $10 and I got 4298 impressions.
Mobidea registered only 489 visits. Thats roughly only 10%.

Am I right to assume that with direct linking the impressions should be about the same as the visits for POP traffic?

When I test the link to the campaign from Mobidea and test it in the browser it definetly works.
I maybe thought it might be because I used https - but actually the link mobidea provides is explicitly https.

Thanks!

Use a different tracker like Voluum or bemob to collect the data.

Good luck!



LeadCloak

Originally Posted by leadcloak

use a different tracker like Voluum or bemob to collect the data.

good luck!

Mobidea offers great added value to their service by offering a tracker for free!
However, it might be a good idea to track your campaigns using another tracker like Bemob which offers a free plan for a limited number of events.

I see you that you have not included any landing pages to your first campaign. Although it may sound a bit more complicated since you will need to rip and clean landers yourself, I would strongly recommend using them for these simple reasons.
- You can increase your offers conversion rate by exciting the visitor on your landing page
- You can split-test different angles as well as designs to see what works best for your traffic
- You have better visibility over the quality of traffic you are buying. Metrics like campaign link visits and landing page ctr will help you a ton on better optimizing your campaigns.

The more you get involved with right from the start, the easier it may get in the long run - at least this is what works for me

My idea was to keep the steps as fast paced as possible. But yes, after you all recommended it - I will go for a external tracker.
I see why you recommend a self-hosted tracker twinaxe, just makes sense with the ammount of events when you scale POP.
However, I don't want to get involved too much in the whole setup and leave room for mistakes in my infrastructure. That's why I go for BeMob now.
But self-hosted makes definetly sense later on when the amount of events drives up the pricing...

Originally Posted by twinaxe

Did you bid $2 fixed CPM for Nigeria traffic?
It´s way too high, better run with lower bid and SmartCPM

Originally Posted by twinaxe

Btw, Herzlich Willlkommen bei STM

Originally Posted by platinum

I see you that you have not included any landing pages to your first campaign. Although it may sound a bit more complicated since you will need to rip and clean landers yourself, I would strongly recommend using them

The more you get involved with right from the start, the easier it may get in the long run - at least this is what works for me

Originally Posted by robertson

Grüß Gott, Servus, Icke oder ehr Moin Moin? ;-)

Today I switched to BeMob and activated my traffic again for a few hundred impressions. No more gap between traffic source and tracker - nice!
Since direct linking doesn't work anymore I jump to using lander in the 40 Day tutorial.

I also signed up with clickdealer - they definetly seem to have a big choice of offer. Now I see the reason for mass-testing offers :-)

Tomorrow is a new day!

And thanks for reaching out to me iAmAttila and alice_cd!

Today I just managed to setup domain, hosting, cdn and signing up for adplexity.
I hope my setup was all correct and my fully active domain is pointing tomorrow to my cdn link.

But I am definetly excited to upload the first view lander and start sending some traffic

-R

I think I can say that I finished to clean up my first lander...

After reading the 40 day tutorial I came up with the following checklist so far.

Lander Fixing:

• Functionallity
• Add missing getUrlParameter Function
• Fix old Date/Time
• Fix Up Outgoing Links to my offer
• Fix BackButton Link
• Device Detection Link
• Fix references to hosted files
• Remove code I don’t understand
• Check for Cookie Setting with cookie-checker
• Check for calling out external services in developer tools -> network tab

Today I played a little with improving lander speed.

You can see in the screenshots the results from every improvement.
To make it short: from 1.97s with 11 requests to 1.22s with 9 requests
Especially moving the start of redering from 1.9s to 0.7s is a big deal!

1) Version as I ripped it


2) After moving style-code from separate .css file into index.html


3) After moving 2 javascript functions and loading of jquery from head to end of file



This part is actually quite fun so far. I enjoy the coding and fixing.
Currently I am a little afraid that I might not be the biggest fan of testing hundreds of offer/geo locations with the landers ..I hope I can also find a efficient way to get through such monotonous tasks :-))

Cheers!

Originally Posted by robertson

Today I played a little with improving lander speed.

You can see in the screenshots the results from every improvement.
To make it short: from 1.97s with 11 requests to 1.22s with 9 requests
Especially moving the start of redering from 1.9s to 0.7s is a big deal!

1) Version as I ripped it


2) After moving style-code from separate .css file into index.html


3) After moving 2 javascript functions and loading of jquery from head to end of file



This part is actually quite fun so far. I enjoy the coding and fixing.
Currently I am a little afraid that I might not be the biggest fan of testing hundreds of offer/geo locations with the landers ..I hope I can also find a efficient way to get through such monotonous tasks :-))

Cheers!

robertson, great job with speed optimization. Although, I think that for a simple lander 9 server requests is still a lot.

Did you use base64 function to load images? If not, I highly recommend you to do it and test again page speed.

Here is the tool you can use: https://www.base64-image.de/

Originally Posted by wakeboarder

robertson, great job with speed optimization. Although, I think that for a simple lander 9 server requests is still a lot.

Did you use base64 function to load images? If not, I highly recommend you to do it and test again page speed.

Here is the tool you can use: https://www.base64-image.de/

I usually use base64 for all small images like: testimonial avatars, logos, small design elements, ...

Very nice start robertson! You've made it through all the boring stuff - the actually traffic and testing using landers will be exciting and fun!

Kudos on switching to a self-hosted tracker and jumping to lessons on landers! I'll be updating the tutorial soon, mainly to incorporate these changes.

Great job on optimizing landers! How many have you ripped and fixed up? I would suggest 5 or more - that will maximize your chances of including a good lander in your split-test.

If you'd like to apply to more affiliate networks, here's a tool that will allow you to do so efficiently:

https://stmforum.com/forum/showthrea...shing-1-button

For initial testing, try to stick with AM-recommended offers that allegedly have good track record (from other affiliates). Once you find a decent lander+offer combo to use as benchmark, you can test any lander or offer you want. But in the beginning, it would be better to stick with proven landers and offers if possible. This post explains it in detail:

https://stmforum.com/forum/showthrea...l=1#post348214

Impressive progress you've made in such a short time! Looking forward to seeing some stats soon!



Amy

Originally Posted by vortex

Very nice start robertson! You've made it through all the boring stuff - the actually traffic and testing using landers will be exciting and fun!

Originally Posted by vortex

Great job on optimizing landers! How many have you ripped and fixed up? I would suggest 5 or more - that will maximize your chances of including a good lander in your split-test.

If you'd like to apply to more affiliate networks, here's a tool that will allow you to do so efficiently:

https://stmforum.com/forum/showthrea...shing-1-button

Originally Posted by vortex

For initial testing, try to stick with AM-recommended offers that allegedly have good track record (from other affiliates). Once you find a decent lander+offer combo to use as benchmark, you can test any lander or offer you want. But in the beginning, it would be better to stick with proven landers and offers if possible. This post explains it in detail:

https://stmforum.com/forum/showthrea...l=1#post348214

Impressive progress you've made in such a short time! Looking forward to seeing some stats soon!



Amy

Today I implemented the bot tester script.
Since I like to be very systematic my idea was to setup a dedicated bot tester LP for my sweep vertical.
As an offer, next to the "bot offer", I use monetizer.
I added this to my tracker and also created a campaign that works as a template for all future bot tests.
This way I just need to copy the campaign, adjust the traffic source and start testing.
..I added two screenshots for you.

This should allow me to test bot traffic fast when I enter a new traffic source without investing much effort and time in the setup of the test.

Last week I was spending some last warm days in Portugal before the German winter kicks in ...

I am now at the point that I have 6 lander cleaned and ready for Iphone sweeps and the bot test is also set to go...

In my head the game plan was easy: Find Iphone sweep offer in Tier 2/3 Geos and test the heck out of my lander and offers.
..what a disappointing moment when I looked at my offer sheet after going through Mobidea, clickdealer and yepads.
I personally see big things coming from Africa in future, so I set my focus on such countries.
In the end I had 2 Kenya and 1 South Africa offer (4 offer applications on clickdealer still pending)
And those 2 Kenya offers aren't even for the same Iphone model..

I applied now also for TORO Advertising and Aragon to get a little more choice.

However, I am also very open for any good advice that helps me find more offers for a fixed set of parameter (Geo, Offer--> Lander).

I will also check Adplexity to get a idea if there are other sweep offers that are currently trending and where I might find more offers of the same kind..

Or is it just not the reality that I find 3-5 offers of the same kind to run a test for a particular GEO?

*just had to add this positive note
I checked Mobidea and saw my first conversion! It came through the smartlink that I used for seting up my bot tester.
Almost green..spent $0,58 on just seeing some impressions coming in and got a $0,42 payout for the conversion.

Cheers!

Just a quick update...

I had tested 2 offer for a phone cleaner app in IQ. Basically the same offer from different AN (with different Payouts).
I tested them also on both available carrier.

Currently I still don't have a winning pre-lander to move forward. I used two for testing ..both in English.

Both offers converted once and each of this conversion came through a different lander. However, the numbers are saying no reasons to let it run any longer...




Next steps I see:
- run same offer with arabic pre-lander
- keep eyes open for new cleaner/antivir offer

Anyway, I feel happy about having the whole hosting/tracker/AN/TS setup running and focusing more on offer and pre-lander now.
Now my goal is to make use my daily testing budget till I find something promising. That's why I added this little column "Days in the Game" to my dashboard - to remind me that I just have to stick to it with consistent work and with a growing number of days it will become likely that I see the results.

...noticed something already several times during lander cleaning - it's time to ask the forum

I see lander that seem to hide some of the lander logic on their server with php.
An example for such code is the following

<script type="text/javascript">
        var scroll = 0;
        window.onscroll = function() {
            if (scroll != 1) {
                scroll = 1;
                var o = document.createElement("img");
                o.src = "https://zmaav.site/click.php?event1=1";
            }
        }

<script>
        function pushToTrackerViaImage() {
            var pushURL = 'https://zmaav.site/click.php?event7=1';
            var img = document.createElement('img');
            img.src = pushURL;
            img.style.display = 'none';
            document.body.appendChild(img);
        }

This is part of the script from Binom tracker.

Using this script you can track people who spend more than X seconds on the pre-lander.

https://docs.binom.org/events.php

Ah! Interesting!

So basically to segment the audience that showed some interest for retargeting purposes?
Is retargeting really a thing for POP campaigns?

Honestly, I'm not using this yet but I would say that marketers are using this more for detecting traffic quality that for retargeting purposes.

Regarding your campaign update for today.

Budget
If your budget allows you to, increase the daily budget. Spend as much as you are comfortable with and test multiple offers.

Offers
There are many approaches on how to start testing sweeps offers, but what worked the best for me was focusing only on one type of sweepstake offers - Samsung/iPhone offers.

This way you will exactly know what you're looking for and you need to take care only for translations. When you have a winning lander, optimized for speed etc. you are just translating it into different languages. This will help you to be organized and fast when testing new offers.

It's important to test many offers because a good offer is converting well right off the bat. And when you find an offer like this just pause everything else and try to optimize it to green numbers.

Just running a campaign and first time I want to try WiFi and Carrier traffic.
Therefore, I did split them in 2 seperate campaigns..

I am now very surprised that I see much less volume for WiFi than for Carrier.
I want to be sure that my setup is right. That's a screenshot of the propellerads WiFi Campaign targeting. Can anyone validate that this is targeting Mobile Phones that are connected to a WiFi network?

https://imgur.com/a/D8jvkx2

Hello,

I currently see this %26 for the ISP parameter.
Does anyone know how this happens and how to clean this up in the pre-lander code?



Thanks!

robertson, copy/paste part of the code here it will be easier.

The easiest way to test dynamic variables on your pre-lander is that you manually type into the URL the value eg. in this case ISP.

Hello,

the code from the lander:

 function contains(b,a){b=b.toLowerCase();a=a.toLowerCase();return b.indexOf(a)!==-1}function getURLParameter(b){var a=decodeURI((RegExp(b+"=(.+?)(&|$)").exec(location.search)||[,null])[1]||"");if(b.indexOf("brand")!==-1&&contains(a,"generic")){a="Phone"}if(b.indexOf("model")!==-1&&contains(a,"android")){a="Phone"}if(b.indexOf("browser")!==-1&&contains(a,"android")){a="Internet Browser"}return a}

<h1>Dear <script>document.write(getURLParameter("isp"));</script> user, congratulations!</h1>

Hello stm.

another day, another interesting finding

I went through my data from the previous day and checked by the hour since I was curious how my CTR was throughout the day.
Very strange is that in the hour 21 suddenly the traffic spiked but even more strange that with it the CTR als drastically jumped up.

I had the campaign running as "distributed" on PropellerAds with SmartCPM ..

Any reasoning why/how this can happen?

Hello stm community,

a few days of learning and experimenting went by...
Kind of established now my lander that I use for testing and focus on testing as much offers as possible...

current stats are the following:


Today, I actually have a very promising ROI of -13% without cutting anything so far.
However, the volume is just miserable.

The campaign is running in HK on PropellerAds.
I have one campaign for all carrier and one for WiFi.
The Traffic forecast for my WiFi campaign looks the following:
I am bidding $0,91, which puts me in the upper region of the impression share

My targeting:


After one day running this campaign I got 4.300 Impressions which somehow doesn't fit with the traffic forecast.

Do I just have the wrong logic how this works or what can I do in this case to get more volume out of this?


Thanks for your help!

One little tip.
On Propeller Ads it doesn't matter if you choose "All", "Mobile" of "Other" as connection type.
The traffic estimator won't reflect the volume for the specific connection types, it only show the overall volume.
To get better traffic estimates select mobile and then select the specific carriers for your Geo.
Then you get more accurate details about the available carrier traffic and with these info you can also calculate a more accurate estimation for Wi-Fi traffic.

hey Robertson, did you ever find networks that are 1) newbie friendly? 2) have good selection of sweeps offers?

please share if so

Originally Posted by twinaxe

One little tip.
On Propeller Ads it doesn't matter if you choose "All", "Mobile" of "Other" as connection type.
The traffic estimator won't reflect the volume for the specific connection types, it only show the overall volume.
To get better traffic estimates select mobile and then select the specific carriers for your Geo.
Then you get more accurate details about the available carrier traffic and with these info you can also calculate a more accurate estimation for Wi-Fi traffic.

Originally Posted by sheet731

hey Robertson, did you ever find networks that are 1) newbie friendly? 2) have good selection of sweeps offers?

please share if so

First of all - many apologies about the delay in giving feedback. Somehow every time I had your follow-along tab opened, I got distracted and forgot to come back - a result of doing too many things at once.

Good thing @wakeboarder and @twinaxe came in to save the day! Many thanks to them!

Will start with your 10-21-2018 post and work forward...although you must have moved on from some of the campaigns, the advice can still benefit you for future campaigns.

Today was the day, finally I took the decision to change my tracker.

I started with BeMob since it was very similar to the almighty Voluum but offered lower rates and for me as a beginner even a free start for the first 100.000 events.
That was all I needed for my start: No initial setup needed, no initial cost.

After I mad my first tests and even saw my first green pennys I realized that the 100.000 events are not cutting it anymore and a quick calculation showed me that tracker cost would cut pretty much into my budget. With this in mind I kept my eyes open for alternatives. Twinaxe was often mentioning Binom as a very good alternative for POP traffic and was actually in his review also very convincing in terms of functions.

Anyhow, today I decided it's time to switch. Even so self-hosted, it didn't took me long. I decided to use a dedicated server from Hetzner with 64GB Ram - about 46€ / Month with hosting in Germany.
Before I pressed the buy button I had a quick chat with the binom support to clarify that the server is a solid choice. ..they confirmed that this is likely enough for +500k clicks per day. Later on I found this link on the support page: https://docs.binom.org/faq.php#6

The server setup took about 30 minutes until I had access via ssh to my first dedicated server. After registering with binom they offered to setup everything for me. However, I felt like I want to learn something and took the challange to enter the 2 commands into the terminal ..2 minutes later everything was done.

I followed the easy to understand support page to setup the first TS + AF + Lander + Campaign and started sending traffic. ...clicks are registered - perfect!

One thing I noticed during this initial setup was the very fast interface. Even so I was sitting in a cafe with the typical troublesome wifi (...Germany) reaction speed was much faster than bemob.

Next thing on my todo: actually making use of my setup for +500k daily clicks :-))

Since I was a little annoyed about all the manual steps when editing and uploading my landers to S3 I decided I have to automate it.

Because editing happens mainly in the editor, my solution is also based on sublime text.

I basically make use of the S3 command line tools and sublimes builder system.
All I have to do now after saving the file with ⌘ + S to run the builder system with ⌘ + B and everything is uploaded and online.

Next I would add all my lander variations I created in one of those spare "creative sessions" to the bulk upload of Binom and I am ready to go

For everyone interested, a copy of my personal notes how to setup the system. (For Mac OS X + Amazon S3 + Sublime):

STEPS


Install aws

pip install awscli


OR check https://openedx.atlassian.net/wiki/s...+Mac+OSX+Linux


Configure aws tools


* In your Amazon web console, click on your name in the top right, and choose "My Security Credentials". Click through to "My Security Credentials".
* Open "Access Keys", and choose "Create New Access Key".
* A sub-window will open informing you that your access key has been created. Choose "Show Access Key" and copy the access key and secret access key to a text window somewhere. Save them if you want a permanent record of them.
* In your Terminal or Command window, type aws configure .
* When prompted, enter your access key and secret access key. Copy-paste is usually the easiest way.
* Hit Enter for the other two values (Default Region Name and Default Output Format).


Create Sublime Build System


in Terminal
cd /Users/rob/Library/Application Support/Sublime Text 3/Packages/User


nano s3up.sublime-build


Write to the file:

{
"cmd": ["aws", "s3", "sync", "/Users/rob/Dropbox/Entrepreneurship/Affiliate/Lan$
}

Save.

Run in Sublime

⌘ + B
or
Tools > Build

Note: I couldn't post this in a separate thread (can anyone help with this?) so I'm leaving it here.
--------------------------------------------------------------------------------------------------------------------------------------------------------
I'm following @vortex's (Amy's) 40-Day Tutorial and one of the sections is on editing and uploading your landers to an AWS S3 Bucket.

One option is to use CrossFTP as described in the tutorial Day 19-25: Fixing Up Landers.

Another option which is a bit quicker, is to automatically have Sublime sync your S3 bucket with your local files.

User @robertson talks about it a bit on this post. I have a background in coding and I'm going to use this thread to elaborate his method a bit further and make it more accessible.

Note: This guide uses Mac OS X + Amazon S3 + Sublime, but you can still make it work on Windows by tweaking some parts here and there.

Requirements

Pip - Pip is a python library manager that we'll use to install AWS's Command Line Interface (CLI). Please make sure you have PIP installed on your machine before moving forward.

How to install Pip on Windows, Mac OS & Linux

Verify that PIP is installed by opening your terminal (mac) OR Git Bash / any other unix terminal (Windows) and typing

pip --version

 pip 18.0 from /usr/local/lib/python2.7/site-packages/pip (python 2.7)

pip install awscli

aws

 
usage: aws [options] <command> <subcommand> [<subcommand> ...] [parameters]To see help text, you can run:

aws help
aws <command> help
aws <command> <subcommand> help
aws: error: the following arguments are required: command

aws configure

cd /Users/anthony/Library/Application\ Support/Sublime\ Text\ 3/Packages/User

s3up.sublime-build

{
"cmd": ["aws", "s3", "sync", "/Users/yourName/landers", "s3://your-bucket-name"]
}

/Users/anthony/Desktop/Business/AFFILIATE/s3

Mine looks like this:

A terminal at the bottom of your Sublime Text should show up saying the following:

• If you delete any files from your local machine, it will NOT delete them from your bucket
• If you modify any files locally and push it to your bucket it WILL overwrite the existing files
• If you find yourself needing to copy files FROM your s3 TO your computer, you can either modify the script in the file by switching the directory and s3 url around, or just use an FTP (which is way easier).

@0192837465 - Wow!! Talk about starting with a bang! That's some generous sharing of knowledge - I'm especially impressed since I'm not technically-oriented!

Please DO start a separate thread and post it there - I could do it but I don't want to get the credit for your hard work.

How is it that you can't start a separate thread?

When you go to this section:

https://stmforum.com/forum/forumdisp...-Follow-Alongs

Are you seeing this button?



If not, please post below and I'll investigate!

When you make that post, I'll link to it from the tutorial - will really help some members out!

Thanks so much Anthony - I look forward to your future posts!



Amy

Originally Posted by 0192837465

Note: I couldn't post this in a separate thread (can anyone help with this?) so I'm leaving it here.
--------------------------------------------------------------------------------------------------------------------------------------------------------

Code:

{
"cmd": ["aws", "s3", "sync", "/Users/yourName/landers", "s3://your-bucket-name"]
}