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

---

Professional Athlete Tries POP - Follow my journey! (34)

---

Hello everybody!
I have been looking forward to this day for so long now. It's the day I start with paid traffic, and it's finally here!

First a short introduction

I'm Alex, a 20-year-old professional cross-country skier from Sweden. My goal is to represent Sweden in the Olympics one day, but the reason I'm here on STM-Forum is to succeed in Affiliate Marketing. I don't know when I found out about affiliate marketing the first time but it was about a year ago that I got really interested and started reading more into it. In the middle of April this year I tried drop shipping after seeing a long video tutorial on youtube. I set up a Shopify website with AliExpress products and monetized it on Facebook. It failed miserably! I was too afraid to spend money on ads that I quit after spending about $60 I think.

Shortly after I made a ski-blog and tried doing simple SEO. On the blog, I was monetizing ski products through banners and links in the blog posts. Once I realized that ranking in Google was hard I tried sending traffic to the blog through an Instagram page. I managed to build it up to almost 1000 followers pretty quick and got a few conversions. I then added a newsletter they could sign up to and continued to try to monetize there. Got about 80 people to sign up to the newsletter and got a few more conversion. But I soon realized that I was only getting 3% payout for the products and that it was going to be hard to make any decent money with this (since it was a limited amount of people I could reach in this niche).

I took a break and focused on training and school during the summer. Then by the end of July, I joined STM-Forum and started studying like a madman! I was reading and taking notes every day, and I was having a blast! I couldn't wait until I knew enough to start. And here I am, ready to start, almost two months later. So let's get started!


Also: I'm looking to network and meet friends so if anyone wants to talk and discuss ideas, please add me on skype (live:alexkarlssonn_1).

------------------

I'm following Amy's 40 Day Pop Tutorial but I'm jumping around a bit and will start using Landing Pages asap.
This is what I have done so far:

-FunnelFlux (managed) as my Tracking System
-PropellerAds as Traffic Source
-Mobidea for offers
-Started an offer spreadsheet and campaign diary.
-Learned how FunnelFlux work and verified that the tracking is working
-My first campaign launched!

Campaign Info:

Offers: 5 low payout offers on Mobidea ($0.21-0.43 payout. Games, Video, Music)
Geo + Carrier: BR + TIM
Budget: $10 (stopped it after it spent the budget)
Bid: $0.74 (I have some question about this below)

Here is some pictures:




As you can see the clicks loss to Mobidea is huge! Why is this?

Also about the bids. If I understood Amy right, you are supoosed to bid above the average? So for this campaign, the average was $0.5. That's why I bid $0.74. But Amy also said that 1.33-2.33 was ideal to bid, and this is what confuses me. Should I have bided higher even though the average bid was about $0.50?
Here is a picture:


I haven't gotten any conversion either, maybe this is due to the low bid? Or is it time to move on and test new offers? (this is what I'm thinking)


I also wonder what is recommended when it comes to user activity? (see picture below)

I just choose "All" since I thought that was the standard.

If I have left out some important info about the campaign just ask and I will provide it.

--------------

I also just launched my 2nd campaign!! Feel great to get some more experience under my belt.
This time I choose India+TATA as Geo+Carrier combo, and I'm testing 4 different offers ($0.13 payout).
I wanted to experiment with the bid (since I was confused about that part), so this time I set the CPM bid to $2.50 (Totalt budget $10 just to test). I was getting very slow traffic (probably since it's night time there) so I will leave the campaign running while I sleep.

-------------------

Questions for today:
-Why is the click loss to Mobidea so huge?
-How should I go about the bids?
-Should I keep testing offers until I find converting ones (so that I can get some experience with optimizing), or should I just move on to Landers right away (This is what I'm thinking is the best idea)?
-Is the reason I didn't get any conversion because of the offers, or could the bid have played in?

Thanks for today, see you guys tomorrow!

---

I love Sweden, and I even did some of my graduate studies there.

I don't do pops so can't help much there, but there are many on the forum who do who I am sure will chime in.

Welcome to the forum, and good luck!.

---

Welcome to the forum.

Click loss can have different reasons, hard to tell without knowing more about your setup.
The 40day tutorial is nice but little bit outdated in some points because direct linking is not that good anymore, Voluum is not really recommended anymore and Propeller Ads also made some changes since the tutorial came out.
So better continue with landers.
Good thing is that you don't use Voluum because it would be too expensive.
I don't work with Funnel Flux myself but there are also very good cheaper solutions for tracker and dedicated servers.
Brazil and India can be pretty tough.
There is lots of volume available but carrier billing offers are mostly rather hard to convert in these Geos.
I would recommend to do some research for better Geos.
About your bid, don't use fixed CPM, use SmartCPM bids instead.
Will save you good money.

---

Thanks for the response.

Didn't know about SmartCPM until now, so I will definitely use that instead.
My 2nd campaign that was running in Inda had way less click loss then my first campaign running in Brazil. So maybe the huge click loss had to do with the offers or geo+carrier? I can provide more info about my setup if that helps, just tell me what info I need to provide.

Also, I will move on to using Landers asap.
Do you have any recommendation for good geos to try? Or any tips on what I'm looking for when I do research for good geos?

---

Hi Alexander!

Welcome to the forum and kudos for the follow along!

Like twinaxe recommended you, start using LPs as soon as possible.

Regarding the good geos and a research... At the beginning look at the geos where payouts are low (eg. below $0.25). That way you will get data faster and you can learn from it, run split-tests etc.

I would look for SOI offers in geos like: ID, HK, TW, KR.

Look for offers with Adplexity (example settings > Last 14 days, received the most traffic, running = 4+ days) and talk with your AMs. Ask them for SOI/DOI offers with the most volume in the past 14 days.

Don't forget to check new offers daily on your fav affiliate networks. To test a SOI/DOI offer in tier2-3 geo you really don't need to spend much. That's why I was always new and unproven offers with smaller budgets. Especially later down the road, it will be much easier when you will have your winning landers translated into multiple languages. You never know when you find an offer that will convert great and you will be on the of the first who will run it.

Keep up posted!

Good luck.

Cheers,
wakeboarder

---

Thanks for your tips Wakeboarder!
I have been reading your POP FollowAlong and it's been a big inspiration.


UPDATE:

I have gotten my FIRST conversion (2 conversions actually) from my 2nd campaign (running 4 Direct Linking Offers with $0.13 payout in Inda+TATA as Geo+Carrier)!

I will obviously not continue with this campaign, but I'm still happy to have gotten my first conversion.

What I have learned:
-Use SmartCPM when bidding next time
-Setting up tracking and postbacks + verify that it works
-Update traffic cost in my Tracker

Questions I still have:
-What is the best way to bid? Should I go by "slightly above average for my particular targeting" or go by Amy's $2.33 regardless of where I'm buying the traffic?
-Why was the click loss so huge in my first campaign running in Brazil?


Making progress on implementing Landers:
I have been working on getting Landing Pages to work the past day and I think I figured it out. I'm using Amazon S3, CloudFront and Route 53 for DNS. I have been following tutorials here on STM for setting this up. I only have a surface level understanding as to what exactly everything does so I'm surprised I got it to work relatively quickly. I can access my Test Lander with www.MyDomain.com/sweeps/iPhone7/en/LP1/index.html but NOT if I remove the www in front. Is this a problem?

I have also applied to 3 more Affiliate Networks using the STM tool, so hopefully, they will accept me.


What's next?
-Implement Landers into my funnel in FunnelFlux
-Get AdPlexity and start ripping Landers
-Learn how to fix up the Landers

When it comes to future campaign, this is my plan:
Focus on only one Geo+Carrier. Split test many offers and Landers at the same time. Once an offer reaches 2 conversions, I will pause the other offers and use that offer to find the best Lander. Then I will re-test the other offers again using the winning Lander (I think it was Amy who recommended this approach).

Anyway, that's it for today. See you guys tomorrow!

---

Hi Alex! Sorry for being late to your follow-along - I've been playing catch-up since coming home from the China Tour.

Thanks for sharing your story. Looking forward to seeing you compete in the Olympics one day! Is your ski blog still live? I think we can talk more about this - I did a quick search in google and found skiing-related affiliate programs that pay a lot better than the 3% you mentioned. It's a passionate niche where people are willing to spend money. I wouldn't give up so prematurely!

Twinaxe has answered most of your questions from the first post (thanks!) but I'll quickly address them anyway. He's right about the tutorial needing to be upgraded - I've been trying to do this since half a year ago but haven't been able to find the time - continuing to try to fit it into my busy schedule. Here are some details:

https://stmforum.com/forum/showthrea...dates-Tracking

Main things that need to be changed about the tutorial (twinaxe has pointed out some of them):

-Direct-linking to 1/2-click direct-billing carrier offers is no longer lucrative. This will still act as a good first exercise because direct-linked campaigns are easy to set up, but you should then proceed to lessons on landing pages ASAP (after 1-2 direct-linked campaigns).

-Voluum has drastically risen their prices recently - I'm planning to rewrite the tutorial for use with funnelflux.

-Changes in user interface (tracker, propeller, Mobidea).

-Propeller has new features/functionalities. (User activity, smart CPM, automatically sending costs to tracker through the {cost} token.)

But don't let the fact that the tutorial is a bit outdated discourage you from following it! Newbies are still finding their first profitable campaigns with this tutorial. Kudos on figuring things out on your own! And of course we'll be here to help make sure you're on the right track!

That IS huge! Are you sure you're comparing stats across the correct date ranges on the tracker vs. Mobidea? The 2 platforms may be set to different timezones.

In Funnelflux you can set the time zone at the top. In Mobidea you can set it in the "Accounts" area. This is where it is:



And then there's PropellerAds stats - and they use EST.

Make sure platforms are set to the same time zone if you want to compare stats between them.


Like I've explained in the tutorial, the $1.33-2.33 are only a suggested "one-size-fits-all" bid range for newbies running their first campaigns. I basically wanted to make sure people aren't bidding so low that they'd be getting low-quality traffic that didn't convert.

If you know how to read the bid charts, feel free to choose your bid accordingly! As long as you're not "scraping the bottom" so to speak, by bidding so low you're getting all the crap traffic nobody wants, you're good.

In the beginning, when you don't have a proven lander or offer, it's important to ensure traffic quality is decent. Otherwise, when you don't get conversions, you won't know whether it's because your landers and offers suck, or if it's because your traffic quality sucks. For a more detailed explanation on this please see:

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

Later on, when you have a promising lander+offer combination, feel free to test different bids - even very low or very high ones. Lower bids will give you lower-quality traffic that convert worse, but it's also cheaper. Higher bids will give you higher-quality traffic that convert better, but it's also more expensive. Either/both can end up being profitable, but that's where the testing comes in.

And, like twinaxe suggested, using Smart CPM at the start is good practice. Smart CPM will Propeller to bid differently for different placements automatically on your behalf, in order to get you traffic from a wide range of placements. This is like ordering the tasting menu at restaurants to sample as many dishes as possible.

If you bid Fixed CPM, you will get traffic from a more limited set of placements, meaning you will likely get little or no traffic from the best-converting, most-expensive placements.

When you run Smart CPM, after a while you'll identify good and bad placements - at that point you can set up Fixed CPM campaigns to bid higher for the better placements for example (to get more traffic from those).


Time to move onto landers!


I would suggest to just choose "All" as well. That feature wasn't there when I was running camps on propellerads, and certainly wasn't there when I wrote the tutorial.

As long as traffic quality is not low, we can use it to adequately test offers and landers.


All answered in previous paragraphs.


Congrats! There's nothing like the thrill of getting your first conversion!

And good call on deciding not to continue - it's time to move onto landers!


Tremendous!

And it's normal to only be able to access the www. version of the pages, because that's the way it was set up. Just make sure to include the www. when you're adding lander urls to the tracker!


Best of luck! I'm adding more and more networks to the tool - so feel free to apply to more networks if you don't get enough approvals.


Hang in there! If you don't know how to code, this part of the learning will be a challenge.

It may help to share the fact that I suck at coding - but the little bit I know is enough for me to get by. You don't need to know a lot. Just go through a couple of beginner's coding tutorials, then start editing landers. After fixing up 10 or 20+ you'll start to get the basics. It gets easier.


Haha yup it's my favorite approach. I like to identify a good lander first because it's the easier factor. Finding a good offer is harder.

The only thing I'd change about your plan: Instead of focusing on one Geo+Carrier, it can be just one geo, because not all offers are limited to one carrier only. Very minor distinction. You have a great plan haha!

Looking forward to your next update!




Amy

---

Well, it's still up and running but I haven't been active on it since early summer. It's cool to hear that you think it has potential. It was only targeting Sweden but I could relaunch it in English. I sent you a PM with more info.

Yes, the time and date are the same. Very strange that I got such a huge click loss... But it seams to only have been that campaign since the other one worked fine.


Thanks for your feedback! It really means a lot
My next update is coming soon...

---

UPDATE:
-Hosting for Landers is working
-Set up a business website to make it easier to get accepted into Affiliate networks
-Applied to set up my own business (only for one person without employees so it doesn't cost anything. If I get accepted I can remove TAX from a lot of things so that would be great!)
-Applied to more networks. Currently accepted by the following: Mobidea, Monetizer, Bitterstrawberry, and Addiliate.
-Bought AdPlexity

I'm really excited to start making campaigns with Landing Pages but I feel like I've hit a wall. So I have a few questions:

Questions:
Why is it so hard to find offers? I'm following the recommendation and are looking for SOI/DOI offers, in tier 2-3 geos with low payout. I want to go into Sweeps as my vertical but most sweeps offer doesn't have a low payout, and they almost never use SOI/DOI. I assume that this means that IF I want to go with Sweeps as my vertical, then it's okay to not use SOI/DOI, even though it's been recommended by many to use it? Right?

Also, if I'm on Mobidea and search for Sweeps, Mainstream for only Tier 2 countries (no tier 3 available) then only 8 offers show up! And these are spread across different Geos aswell. And this pattern is true for all the networks so far. I don't see how I'm supposed to test MANY offers when there is so few of them available. If I'm going to follow the advice of only use one GEO to start with, then it's even less offers available. Does this mean that Sweeps is a bad vertical to start out with (since I want to test offers with low payouts and in tier 3 geos)?
I really want to be able to test MANY offers and really split test my way to victory, but it seems like there is just so few offers. I feel like I have missed something here, and/or are doing something wrong. As for Geos: right now I have been trying to choose Hong Kong, India, Malaysia, Taiwan or Thailand as my Geo to use.

This brings me to my next question. If I instead go with Mobile games/apps or video offers as my vertical (like the ones I was direct linking in my test campaigns). How do I go about finding matching Landers for those offers? I feel like the landers need to be very specific to the offers in those cases, and maybe then it's hard to mass test offers? Or am I wrong here?


Hopefully, I don't sound too much like a complete noob lol. I really feel like I have missed something since I feel like everyone is talking about testing offers left and right. Would really appreciate some help here.

-Alex

---

Don't worry about sounding like a newbie.

Sweeps offers in IN, HK, MY, TW and TH with SOI/DOI offers and payout starting from 3 cents.
https://www.affplus.com/?sort=price_...%5D=Sweepstake



LeadCloak

---

Thank you LeadCloak. Great website you linked there!
New mission: Apply to MANY networks to get as many offers to use as possible. I thought it would be easy peasy to find 20-30 offers for the vertical+geo you had chosen and then just start testing. Turns out it's much harder to find many offers then I thought.

I have made a spreadsheet writing down Sweep offers for South Africa and Malaysia. With the networks I'm currently with I have written down 4 offers each. My goal now is to apply to as many networks I can and get 10 offers for each of my two geo+vertical combos. Then I will start testing them (and laders ofc). If anyone has tips on a better vertical+geo combo that is easier to find large amounts of POP offers (low payout),please tell me and I will try it after my current project.


As for my regular life (if anyone is interested), tomorrow I will be going on a training camp. We will be training inside a tunnel that has snow (it's always snow inside the tunnel, even if it's summer othside). So many hours of skiing awaits me to next couple of days. Will try to squees in as much work on affiliate marketing as possible in between workouts.

Stay tuned guys!

---

Sounds like a lot of fun!

When applying for more networks, this tool can help :

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

---

Replied to your PM.


My mind is still not at ease - have never seen clickloss of this magnitude.

For your other campaigns, what's the % clickloss?


It IS true that you'll need to apply to multiple affiliate networks if you want to test many offers.

But then, applying to TOO many networks will result in a couple of unpleasant consequences:

1)Your revenue will be scattered across lots of networks, all under the payment threshold so end up getting "stuck", resulting in cashflow issues.

2)You won't like having lots of AMs pushing you to drive traffic to their offers.

Therefore, be selective if you can.


Adplexity and offervault are your friend. So are the posts on this forum. Feel free to do one or more of the following in isolation or in combination:


1)Search this forum for posts where the best networks for sweeps offers are mentioned, and apply to them. Some example threads:

https://stmforum.com/forum/showthrea...-in-march-2018

https://stmforum.com/forum/showthrea...rce-in-germany

https://stmforum.com/forum/showthrea...ike-affluentco

https://stmforum.com/forum/showthrea...pstakes-for-fr

Then you can speak with the AMs to find out which geos and offers are doing well and start testing from there.


2)Go to Adplexity. Define a recent date range (a week should do), and select "Received most traffic" from the top dropdown. Search for sweepstakes-related keywords like "win", "iphone", "samsung", "ps4", "voucher", "gift", etc. (and browse the results to write down more keywords that are relevant to sweeps - including keywords in different languages!) Browse the top entries (say 50?) to see what kinds of offers are available in general, and which countries each type of offer is in.

For example when I search for the keyword "iphone", then scroll down to the sidebar's "Country" section, I see this:



So you know which countries have the most iphone offers. If you're a newbie, you may want to go down the list and pick out the tier3/4 geos.

Also scroll to the sidebar's "Affiliate Network" section to see which networks have lots of iphone offers:



Of course the above is just a general example. You can narrow down your search to just "iphone x" or "galaxy s9" for example, to see what the major geos are, then select some of the tier3/4 geo in the "Country" filter, and scroll to the "Affiliate Network" section to see which networks have those offers, and apply to them. (Note: Results aren't 100% accurate though, because "iphone x" will bring up landers that are calling out on the page, visitors that were using iphone x. But you'll get a good idea using this method.)


3)Go to offervault.com. Again, type in keywords associated with sweepstakes offers. Note which affiliate networks and geos have those offers, etc. etc. (similar to what I talked about for Adplexity above).

Sometimes you just won't find "many" offers for a specific geo+prize. Doesn't mean you can't still test them. It would just mean you'll need to spend time preparing all the landers for that prize + language, and take greater risk that you won't find a profitable offer.


It's not easy to find landers for such offers. Please see my explanation in this post:

https://stmforum.com/forum/showthrea...l=1#post345282

In the same post I've linked to some tips you can try, if you really want to make those offers work using landers.


On the contrary, your questions are SPECIFIC which really helps us to help you!


Sounds like fun!!

We'll be here when you get back.



Amy

---

Hi everyone! I'm back from the training camp
Did not have time to do much work during the camp other than learning some basics on how to code (I have zero experience from before). But now I'm back and ready to start hustling!

Thanks for the great tips Amy and matuloo!

Here are screenshots of the direct linking campaigns I was testing:




My first campaign in Brazil dropped from 13 577 (PropellerAds), to 13 162 (FunnelFlux), to 5127! Lost over 8000 impressions
My second campaign in India dropped from 4000 (PropellerAds, to 3638 (FunnelFlux), to 2956. Lost over 1000 impressions (still seems like a lot to me)

--------------
UPDATE:
I choose to go with South Africa and iPhone/Samsung sweeps for my next campaign. I have found 7 offers (spread over 3 different networks). And I plan on getting 10 fixed landers before launching this next test campaign. Your tips really helped me here.

I have started downloading Landers from Adplexity. Right now I'm in the process of fixing and optimizing them for speed. The first one was very tricky since I couldn't find the links that were hidden, so every time I uploaded it to my bucket it redirected to his own offer/landing page. After a few hours of messing around, I managed to remove all the hidden links and add my own. I also fixed up some of the broken stuff and I felt very accomplished when I was done

I fixed up another one and then I remembered that I needed to optimize them for speed. So I compressed the images, moved the java scripts to the bottom and even removed some chunks from the lander that I thought wasn't necessary. I even removed the custom font to speed it up even more. I then ran some speed test at the "Webpagetest" website and I was very disappointed with the results.
I managed to get my first lander down to 1.8s and the 2nd one to 3.0s.
I even ran Amy's test lander (used in the POP tutorial) and while Amy got about 0.27s in her screenshot, it took me 0.87 for the same lander.

I think the problem lies with the "First Byte" load time being REALLY high. I have provided screenshots below.
Worth mentioning is that I sometimes get an F on the "First Byte Time" and sometimes a C. The total loading times has even increased since I did the tests in the beginning, compared to now (a few hours later).

Amy's Test Lander:


My First Lander:


My 2nd Lander:


I would really appreciate some tips on how to improve my loading times. Keep in mind that I have done most of the things recommended already in order to increase loading speed.
Some additional info: I'm using Amazon S3 + CloudFront + Route53 for my CDN setup.

Update(05 October): I got multiple errors from Amazon S3 shortly after my post. The errors were about "Delayed Data Delivery" and Amazon is currently trying to fix them. Maybe once it's solved it will speed up my loading speed. I will post an update here once the issue has been fixed.

---

Update:
The Amazon errors were fixed but the Landing Pages are still loading slow. Over 2s even though I did my best to optimize them. The "First Byte Time" is also still slow at about 0.8s. Anyone who knows why this is?

---

Losing 8k impressions is ridiculous. The 1k impressions is still a lot - although still marginally-acceptable and is something I've seen to happen under "normal" circumstances.

The clickloss from Propeller to FunnelFlux is reasonable. I would examine more closely, the redirection from Funnelflux to the offers.

Are you using the managed or self-hosted version of Funnelflux? If managed, you should definitely ask support to look into your set up to make sure there's nothing that's causing errors in redirection, or the redirection speed to be excessively slow.

If you're using self-hosted - please let me know and we can do some troubleshooting ourselves.


Haha! As a tech/coding-challenged person, I completely get what you mean!


It's been a few days - how's that coming along?

ZA isn't a geo I'd normally suggest for newbies to tackle, because costs are high and traffic volume is massive, which means you may need to invest significant money into testing offers+landers and cutting placements before you can see profits (IF!). But if you can conquer this geo, rewards can be significant too.

Basically you'll need to focus on testing landers and offers until you can make a large-enough portion of the total traffic convert profitably, then cut the rest. And even while testing landers+offers, temporarily blacklist placements aggressively (e.g. at 0.5-1x payout without converting). Something like theoptimizer will do this efficiently for you.

Once you're profitable, you can always retest some of the previously-blacklisted placements to see if they'll be profitable on the best offer+lander.

To reiterate though, don't expect to see profits until after a significant investment! Not saying it isn't possible to see profits right away, but that would be more unlikely than for tier 3/4 geos.


I would suggest to go through the tutorial again to thoroughly check every step of your setup.

Also: You can PM me the urls of your landers, and I'll take a look. Please also zip your lander files and upload them somewhere and let me know the download links. Alternatively, you can make a separate post to see if any coding experts here can give advice on speeding up loading times. (You don't need to post your lander urls in public - either post the download links to the zipped files, or paste the lander code in pastebin and post the pastebin url.)

You can also try these tools that will give you insight on how to speed up loading times:

https://developers.google.com/speed/

https://gtmetrix.com/

Hope that helps!


Amy

---

I'm using the managed version and I have asked the FunnelFlux team for help.

I'm been moving forward so slow in the past days. I have had a lot to do with multiple workouts each day and school, but when I get down to working on affiliate marketing I have been procrastinating or felt stuck when it comes to improving the speed of the landers. But I'm going to pick up momentum now and start launching campaigns.

Okay, then I will probably try another geo right after this. I will however finish and laucnh this campaign since I already have started with ZA. But I will move on to another geo right after since you seem to think it's wise.

Yes I will try that. Will upload the landers to Dropbox and PM you in a few minutes.

---

UPDATE:

First campaign with Landers launched (finally)!!!
I'm testing 7 offers from 3 different networks and 7 Landers (used to be 11, but 4 was not approved by my AM from one of the networks).


I added the bot-test and hopefully, it will work. EDIT: Can anyone explain how I view the stats for my bot-test? I want to see the CTR to my Pares Javascript offer, from the different placements. This way I can cut the placements with high& bot-traffic.
Also, checked tracking for myself before and everything seemed to work fine. I just launched the campaign and so far everything is working as it should.

The problem from before with slow load speed is still there so any help on that matter would be great.


Regarding the previous click loss:
Zeno said that it was either because Mobidea only shows unique impressions, or because the loading speed was really slow on those offers. I checked my stats in FunnelFlux and thinks it's a mix of both.

Next goals:
Start launching more campaigns!
I want to have a system so that I can easily create multiple campaigns each week. So my first objective is 1 campaign each week, then 3/week. And hopefully I can launch 5/week or even 1/day once I get the hang of it
Some feedback is always appreciated.

---

Hello Alexander!
Thanks for the nice post, concrete questions and answers from the experts!
I'm in similar situation, starting ripping and adapting landers, picking offers, installed Binom, domain, digitalocean, namecheap, ...
What I don't understand is the technical part of dynamically linking and rotating LPs and the offers:
- I have the LP url and a CTA button on it with a fixed path to the offer.
- I have the Offer URL.
I mean if I have 10 LPs and 10 offers I need to create 100 offer combinations or Binom manages these connections by cookies or whatever?
Thanks!

---

Hi nviktore.
On the LP you should not have the offer URL on your Call to Action button. You should instead use your trackers Action URL which will then send the visitors to the offer/s. In FunnelFlux for example, the action url takes the visitors to a rotator which split-tests the offers. I don't know the technical details for Binom but I'm sure there are tutorials out there.

Hope that makes sense.

---

Quick update:
I found out that the Bot-Test only works for (1 Lander+1 Offer). That could explain why I couldn't figure out how to view the results from the Bot-Test in FunnelFlux.
So I decided to remove the Bot-Test offer. Is there another way to have a bot-test at the same time as you are testing multiple landers and offer? Or do you need to find a successful lander+offer comb first, and then do the bot-test after (or maybe run the bot-test first and then blacklist the bad placements, and then do the landers&offers split-tests)?

---

Hi Alexander! I am no expert (and I have not yet done bot testing myself, in progress), but I see you took the implementation from here:
https://stmforum.com/forum/showthrea...live-campaigns

Logically it shouldn`t matter how many offer-lander combos you are testing simultaneously. When Traffic source sends traffic to your campaign URL, FunnelFlux automatically rotates the landers that user end up at. For each of these landers you have set up the bot test part. When a bot (which usually does not have javascript enabled) arrives on your lander it fails to parse the URL added in the bot script part and no clickthrough to dummy "Parses Javascript" offer is registered. When a real user arrives on the same lander, the browser is able to parse the URL and a CTR is registered to the bot detection URL.

Later when you view CTR for placements, you can still see what the percentage for the whole campaign is. It does not matter, that the numbers are aggregated from all the landers in your funnel (with the assumption that all of your landers had the bot script added)

But I might be missing something, since like I said, I have yet to test it myself.

---

Yeah, this is what I thought too. But I couldn't figure out how to view the stats in FunnelFlux. If anyone knows how to do this and could explain that would be awesome!

EDIT:
I think I figured out how to view the CTR for the bot-traffic offer.

Did I do it right? The LP CTR is the amount of NON-Bot-Traffic?
Obviously, those placements have too few impressions to be cut.

Some placements have over 100% CTR, why is this? And has this something todo with bot-traffic?

---

60% CTR to Parser offer would mean 60% human traffic.

Does CTR of over 100 happen to all the landers? CTR of over 100 could happen if for example, you redirect backbutton clicks back to the same lander (which reloads the lander + bot script part)

---

I see... I have the Bot-Test script set up in all my Landers. They link to the Javascript offer. Is there a way to make it not reload the lander?


Sent from my iPhone using STM Forums mobile app

---

This is OK, that you have the bot-test script added to all of your landers. It does not reload the lander, but it`s rather the opposite way - each time you load the lander, the bot script is executed and and a clickthrough to the "parser offer" is recorded. Review your lander - is there some action that might cause reload of the whole lander?

Tomorrow I will also be testing the same bot script - I will let you know about my results and if I see something similar or not.

---

When you say the cost of testing will be more for larger countries, do you mean you spend more than 10-15x payout? If this is the case, how much more would you suggest testing with?


Sent from my iPhone using STM Forums mobile app

---

No, I don't think there is anything that re-loads the Landers.
It's not the CTR to the lander that's over 100%. It's when I drill down into Campaign->Landers->Offers->Zone(Placements) (see pictures in one of the posts above). Then it says that many placements have over 100% CTR. Is this normal? I thought this was the way you were supposed to view the CTR of the "Javascript Bot-Offer". Or am I doing it wrong?

On a side note: I asked the FunnelFlux support team for some help and Zeno suggested that I tried to cut bad placements using a "time on-page script" and then analyze the time the visitors spend on the LP. Low time would mean bad traffic or bots. I will probably try this instead for my next campaign. Any thoughts?

CAMPAIGN UPDATE:
My campaign has spent it's budget and is not stopped ($25, I used Smart CPM and my bid was $2.4), and here are the results (I didn't know how much stats was necessary to show but I think this should be enough):






Only one conversion so I will probably kill this campaign and try a new one. All the low payout offers have almost spent 10x it's payout and is therefore ready to be cut. I could keep running the higher payout offer but I feel like a better investment would be to start a new campaign with lower payout to learn from my mistakes quicker.

Click/view-Loss:
My total impressions from PropellerAds was 19 853. And on FunnelFlux it was 18 568. This seems pretty normal to me. My campaign ran in ZA and my FunnelFlux managed server is in Europe, so I assume click loss could be less if the server was closer to the targeted geo, correct?

I got no offer views registered at my 2 offers at Toro Advertising so something went wrong there, I'm looking into it and because of this, I might re-test the offers from Toro.

However, the offer to Addiliate had a HUGE CLICK LOSS! I got 523 views (488 unique) recorded in FunnelFlux, as seen on the pictures above but only 128 clicks on Addiliate.

Any help here? If I can solve this issue I will probably also re-test this offer with the two from Toro.


On Mobidea, I think the click loss was normal for the first 2 offers, but to much for the other 2. My thought is that it might be because one of those offers only accepted Phone-traffic and I didn't buy traffic for JUST phone. So maybe all the non-phone traffic go filtered out? Any thoughts?




Questions I still have:
-How to improve the speed of landers (see previous posts)
-Why was the click loss to Addiliate (and 2 of the Mobidea offers) so huge?
-Still confused about the bot-test
-Is it okay to find a good lander+offer combo FIRST and then do the bot-test after? Or should you cut bad placements right from the start?

Lessons learned & upcoming tasks:
-Test so that tracking is working correctly for ALL the offers beforehand
-Choose a low tier country next time
-Choose fewer offers (3-5) so I can launch a campaign quicker
-Try cutting placements using on-page time instead of the bot-test as Zeno suggested (or maybe use both)

---

Hi there, Alexander. Regarding the CTR being over 100 - when I chose your filter (Campaign + Lander And Offer + Zone) I had similar results. It seems to aggregate click actions there.

The closest I currently got to detecting bot traffic per zone was:

I contacted FunnelFlux support, to ask if there is a better way. Will probably also look into some other bot testing methods too

---

I finally had a chance to look at your landers!

I've checked loading times across webpagetest, gtmetrix, and google pagespeed insights (tried to use pingdom but it wasn't cooperating - kept getting errors so I gave up on that one):



I think that webpagetest is no longer a good tool for assessing page load speed. Perhaps someone more technically-inclined can analyze the waterfall and see WHY it's giving results that are so different from other tools.

In addition to these tools, I've also quickly check a couple of the landers using the other tools recommended here:

https://geekflare.com/test-your-website-load-time/

...results of which are on-par with those from gtmetrix.

So - congrats and apologies! Congrats on having done a good job on speed optimization. And apologies that I was still recommending webpagetest as the go-to tool when it no longer accurately assesses load times. I'll be sure to rectify that ASAP when I update the tutorial (this weekend!)


On a side note: I asked the FunnelFlux support team for some help and Zeno suggested that I tried to cut bad placements using a "time on-page script" and then analyze the time the visitors spend on the LP. Low time would mean bad traffic or bots. I will probably try this instead for my next campaign. Any thoughts?

Good goal! Once you get the hang of optimizing campaigns, testing widely (but intelligently) would be the best way to find gems to scale.

But I would suggest not to juggle too many campaigns now - you have yet to get more experience on how to optimize and scale campaigns. Juggling too many may just distract and make you drop the ball somewhere. But if you feel you can handle it then don't let me stop you!


Regarding running the bot test with multiple offers and landers - please see this post and the 2 below it where matuloo and I replied:

https://stmforum.com/forum/showthrea...l=1#post339976

Note that everything in those posts were in reference to Voluum. I don't know whether it'd be possible to do a bot test on multiple landers and offers using Funnelflux. Will test this myself when I adapt this tutorial to Funnelflux (scheduled to be done in the next few weeks).

For Voluum (and any other cases where it's not possible to run this bot test on multiple landers and offers), I would actually suggest to run the bot test at the start using any one offer and one lander. Set a minimum campaign budget, then keep refreshing tracker stats and pausing placements/zones that have received 60+ impressions (i.e. enough to gauge bot %). I would keep doing that - until I have to wait around a minute or more without having to pause any further placements, then stop the bot test. That should already give the biggest placements enough impressions (i.e. 60).

Or you can use theoptimizer to do this:



...and just keep refreshing stats yourself until you don't see additional placements hitting the 60-impression threshold for a minute or more. Or if you don't want to keep refreshing stats, you can always set an alarm for 5/10/15 minutes and check back.

Then after that bot test is over, you can run the campaign by rotating all landers and offers as per your original plan.

Doing the bot test after finding a good lander+offer combo would be another approach, but that would be less efficient in terms of saving costs. It's better to eliminate the bot placements first without wasting money on them while testing landers and offers. You're not obligated to run a bot test, but if you're planning to do this, it would achieve the most cost-savings if run at the beginning.


Correct as far as I can tell!


I mean something else entirely. When a geo is big and has lots of placements you'll need to invest more money into cutting unprofitable placements (because there are more placements you'll have to spend 2x or more payout on, to find out whether you need to cut it or not).


Zeno is a master when it comes to all things tech. I would suggest to ask him for more details on how to implement such a test.


Am I reading correctly that the payout is only $0.20? How could that be? ZA sweeps offers have higher payouts than that certainly?

I would suggest to test another english-speaking geo so you can re-use these landers. $1-3 payouts would be good to test without breaking the bank.

Alternatively, if you could find more similar offers for this geo that are suggested by AMs, I'd give that a shot first, since you already have campaign and tracking etc. all set up.


I think I know the answer, but best thing to do would be ask Funnelflux support to confirm.

Either way, I wouldn't worry too much about this at this point - AFAIK, the difference wouldn't be large enough to make/break the campaign.


Yes - something does seem wrong!

The tracker did redirect the visitors to the offer link, so if no views are registered at Toro for the offers, then either:

a)The aff link wasn't set up correctly at the tracker, or

b)Toro has somehow redirected the traffic to other offers - perhaps because the traffic that was sent, was not accepted by the offers. Did you see views show up for any other offer in Toro stats? Best to check with your AM.


Sounds like it's a problem with the addiliate offer. If the offer page loads slow for example, and visitors are bailing, then that's one scenario that COULD be causing clickloss like that. Often it's not the affiliate network but the advertiser (i.e. owner of the offer) that hasn't optimized their offer page for speed. I would test another offer unless it looks really promising in spite of the large clickloss (if that's the case, it may be worth asking your AM to have a conversation with the advertiser, but that doesn't look like it's the case here).


Entirely possible! But if the traffic got redirected to another offer, then you should see stats show up on your aff network dashboard for one or more offers that you haven't promoted directly.


Hope that helps! Please feel free to ask further questions. Looking forward to seeing stats from your next campaign!



Amy

---

UPDATE

So I have started to re-test my TW campaigns. I'm using the one winning lander that I used from before (translated) and I'm testing 2 offers. Today I did 2 test where I tested 2 different bids at different times.

TEST 1
I used the same bid as I had success with before (SmartCPM $0.5) and the same blacklist as well (about 30+ blacklisted zones)
RESULTS
Spent $3.70 and got 3 conversions (payout $0.30) which made it -73% ROI.
All conversions came from Computer traffic (WiFi and Wired connection) and all conversions also came from only Windows as the OS.

TEST 2 (started once test 1 was finished)
Used the same campaign but changed the bid to $1.50
RESULTS
Spent the entire daily budget ($6.79 since the start of this test and $10.09 in total) and got 5 more conversions which made it -78% ROI for this test (so about the same as for the other bid. Total -76% ROI for this day)

4 of the conversions came from Smartphone Traffic, and 1 from Computer traffic. So it seems like my bid change made it so that the traffic targeting changed a bit.

-----

Here is a screenshot (total stats for the day)


So as you can see even when I got only Smartphone traffic after I increased the bid, I still got it from WiFi traffic (and Wired apparently? Can you use wired traffic on mobile?).
And during my original test where I only got computer conversion, all of them was from WiFi/wired (except one which was mobile, I assume this mean they shared their 3g to their computer).

The ROI for Smartphone (and WiFi/Wired) is much better compared to the rest so it might be worth to just target Smartphone and Wifi. I will also try an even higher bid tomorrow and see how the results are.


Edit:
I also found this:


It seems like Chrome as the Browser for Smartphone traffic it doing really well. Might be worth doing Smartphone+Chrome+WiFi then.
Although I feel like it's important to keep in mind that the traffic for that specific targeting is only 15% of my total traffic from today (3600 impressions of 24500 in total).

When doing very specific targeting like this, is it best to do one factor at the time? Like, first just Smartphone traffic, then Chrome etc. Or can you just test them all at the same time?

---

UPDATE December 7

Okay, so a few things have happened since last time.
I have been doing a lot of testing on my TW campaign and I can't seem to get it under -70% anymore. What I still want to do is translate some more landers and test them, and I also want to test some more offers before I give up. I have applied for a few more offers but not yet been approved, so that will have to wait.


I have been testing some campaigns in Spanish speaking geos and tried to test a few offers in Colombia. I had an offer from ClickDealer that sent ALL traffic to a Smartlink instead of the actual offer, even though the traffic requirements matched my settings in PropellerAds. I even tried to create 2 different campaigns in PropellerAds and the mobile campaign still sent all the traffic to the SmartLink offer. But the other campaign didn't send any traffic anywhere?? Not even to the smartlink.
My AM is not coming back to work until december 10th but I have tried contacting clickdealer about this. Has anyone else had a similar problem or know what might have caused this? I have triple checked the tracking links and everything is looking good.
UPDATE: I have found the cause of the problem. This error was due to the "Global Conversion Cap" being used up. That offer could only get 50 conversions per day in total for all affiliates.


Now here is the positive news! The SmartLink is actually doing VERY well. It was actually doing PROFIT during the last 2 hours so I decided to keep running it and try to scale it up some xD

Here are some stats from the past 2 hours:


At first, it was doing between -30% ROI, but then I noticed that like >90% of conversions were coming from one of the major placements. So I just whitelisted that one, and then BAM! PROFIT! This was not the way I expected it to go when getting my first green campaign, lol!

Anyway, I will play around with this a bit more and continue to test new offers and geos.

See you next time!


EDIT (dec 9):
So I have been making between $2-2.5 per day now! It's not much but now I'm actually doing profit! I can't believe it!
Feels kinda cheap though, to be honest, to have found profit this way since it's a smart link that redirects traffic. But I'm still very happy about this! This motivates me to keep testing more geos, offers and landers!

---

Yup if you have multiple campaigns with the same targeting, then the "stealing" of traffic could very well happen.

As for other factors that can cause volatility in performance, some of them are: Variation in visitor behavior throughout the day/week (depending on whether they're on the move, in the office, at home relaxing, etc.), variations in competition levels (for example from other affiliates setting up dayparting), technical issues with the offer (either at the aff network or advertiser)...there must be others, but I've listed the ones off the top of my head as far as volatility goes.

Also: When it comes to a drop in performance in general, saturation can be an issue (especially in small geos). Also, if your lander is aggressive, some webmasters (i.e. owners of placements) would stop you from showing your ads - just like we can cut placements that don't convert, they can cut ads they don't want shown to their visitors.

Lastly: A decrease in conversion rates is normal for pop campaigns - I've summarized some of the things you can do to try to revive the campaign:

https://stmforum.com/forum/showthrea...formance-dives


If the placement is really high-volume, AND your ROI is at least close to breaking even, then it MAY be worth it to test that placement at a lower CPM. Otherwise, I wouldn't bother.


That can be quite a dilemma. I hate it when all the placements that have made conversions are small - with 1 conversion each.

And that happens all too often I'm afraid.

Sometimes you'll just have to bite the bullet and keep dumping money into the campaign to gradually cut down on placements. But before you invest money into cutting, invest money into split-testing offers and landers until you've found a reasonably well performing combination.

Another thing to keep in mind: The conversion rate of a placement doesn't tend to change significantly with the bid. Of course there will always be exceptions, but that assumption will more often than not stand you in good stead. This will help you to decide how to bid also (i.e. you can assume the conversion rate to be constant, and figure out how much to bid in order to make a placement profitable ; keep in mind though that when you bid lower, you'll get less traffic as well).


This is interesting!

I know I often suggest to target big traffic segments separately, but it's also true that the broader you target, the more priority some traffic networks will give to your campaign. What can happen is that you may get more volume and better quality traffic for cheaper by targeting broad. Definitely a good thing to test!

You can also set up tracker rules to direct traffic from different segments to different offers, and try to optimize the performance of every major segment by split-testing landers+offers.


Interesting! I haven't used funnelflux enough to know this, but will find out and get back to you!


[QUOTE]It seems like Chrome as the Browser for Smartphone traffic it doing really well. Might be worth doing Smartphone+Chrome+WiFi then.
Although I feel like it's important to keep in mind that the traffic for that specific targeting is only 15% of my total traffic from today (3600 impressions of 24500 in total).

When doing very specific targeting like this, is it best to do one factor at the time? Like, first just Smartphone traffic, then Chrome etc. Or can you just test them all at the same time?/QUOTE]

Definitely worth a test!

And if the ROI is good, you may be able to bid higher to get more volume.

By testing one factor at a time, do you mean set up separate campaigns to target each segment? That wouldn't be necessary - to see the performance of separate segments you can always drill into your stats.

Also, when you're targeting too narrow, some networks may not give your campaign a lot of traffic.


So glad to hear! And yes a lot of the best offers nowadays have caps - did the offer description not state this? If not then it should! All's well that ends well I guess! But for the future, if you're seeing from your aff network stats that the traffic isn't going to the intended offer, it would be good to put the camp on pause until you can get in touch with your AM.


Running smartlinks is nothing to be ashamed about!

And congrats on your first green campaign!!

If you're really wanting to give that smartlink a good testing:

1)Make a separate campaign with that profitable placement blacklisted, cut placements at a certain multiple of the payout (0.5-2x depending on your style and how quickly you want to limit your losses). When you find more profitable placements, blacklist it in this camp and "transfer" it to the profitable whitelist camp.

2)Make yet another campaign, and target all countries with the lowest bid allowed on that traffic source. See which countries are profitable, and target those separately in new campaigns and test bids and cut placements etc. Cut geos that are spending money without converting. Note that you don't need to spend much on tier 3/4 geos because payouts are generally lower, but for tier 1/2 geos you'll need to spend more to gauge promise because payouts tend to be higher. Also, at such a low bid, it will take longer to get enough traffic to tier 1/2 geos. (For this reason, you may want to target tier 1/2 vs. tier 3/4 geos separately from the start, and set a higher bid for the tier 1/2 camp. Or, you could target all geos in the same camp and cut unpromising tier 3/4 geos first, and keep running the tier 1/2 geos further.)


Have fun!



Amy

---

BTW - just got a reply from Funnelflux regarding the "wired" traffic:

I've taken the liberty of putting the important part in bold. Hope it helps!



Amy

---

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