Home > > Newbie Follow-Alongs

---

Yet Another Pop Traffic Follow Along (9)

---

Hey everyone, so after getting stuck in a loop of continual reading I decided to just get going again -- and I thought, why not create a follow along? :-)

Quick history: So far, beginning at the tail end of last year, my AM experience has been launching a handful of pop traffic campaigns with some limited success (a few conversions here and there). However, in retrospect, I probably bailed too early on a number of these.

My objective here is simply to further develop and improve my approach to AM (using vortex's tutorial as a loose guide).

I plan to once again use pop traffic (e.g. Propeller Ads) and I have no particular vertical in mind. Since I have no idea what works I'm going to have to rely on external data from an affiliate manager.

So, next phase: find some offers!

Thanks for taking the time to read this, I'll have updates (and questions) to come!

---

Good luck with your follow along!

A few suggestions:

1) Split test direct offers with Monetizer / Afflow to hedge your bets.
2) Start with Propeller and another source ideally. I would say ZP is good for beginners.
3) Use a spy tool like AdPlexity so you don't start blind

Let us know how it goes

---

Hi realtalk and thanks for giving my tutorial a try!

I'll be around to answer your questions.



Amy

---

First, thank you bbrock32 and vortex for dropping in on the thread

Regarding suggestion #1, if I understand correctly, Monetizer is a service which can help monetize unused traffic (e.g. traffic I do not send directly to an offer or lander of my own)? Does this assume that I am buying untargeted traffic and doing some routing at the tracker level for each offer? I haven't done this before but you got me thinking about it now!

In any case, I made an account with them and am still awaiting approval for the time being.

---

Monetizer will send your traffic to the best combo LP / Offer based on many factors like country / connection type / device type etc.

So if you buy Italian traffic and send to monetizer, this traffic will go to the best LP / Offer they have for Italy.

It's a good benchmark to compare with the offer you want to run through a network.

---

Hey guys, no, I did not abandon this thread entirely! :-)

Quick summary: This was a pops + sweeps campaign where I ran 3 landers and 2 offers (ended up getting whittled down from 5 to 2 after I discovered 3 of the offers were actually dead and still listed as active on the affiliate network!)

The campaign ROI was -95.58% so I decided to just cut my losses on this one



For landers I ran a few questionnaire style ones where the user answers some questions, ends up on a list of offers to choose from (for having completed the "survey") and goes from there.

My experience with ripping landers was finding out that it can make a BIG difference which ones you pick from Adplexity. Some are easy to fix up, and some have the most convoluted logic inside of them ;-) So make sure to glance at a few varieties of the lander code before you invest time fixing it up (especially if doing it yourself!).

As for browser breakdown, I noticed it come up in a few threads now, the occurrence of "Android Webview" traffic was new to me.



I found it interesting that my lone conversion came from a PS4 of all places!

I just checked however and it appears Propeller Ads will let you filter it out if you need to.


Here's a brain dump of decisions I made that caused me to spend far more time launching the campaign:

* I chose offers myself instead of going strictly with suggestions from AMs (note to self: get on with more networks, it was difficult finding a group of similar offers in a single GEO)
* I picked a non-english GEO (Italy in this case!) and had to translate my landers.
* I saw the news about the Chrome HTTPS update and then changed my hosting infrastructure (i.e. get certificate for CDN, getting HTTPS tracking working on Voluum)
* Related to this, for Aff. networks that allow you to use custom tracking domains, you may need to use their domains instead of your custom ones if you want to have HTTPS all the way through the funnel (tracker -> lander -> AN offer link).
* Went down the rabbit hole over-optimizing landing pages and also creating one of my own from scratch (and it performed no better than the ripped varieties!)
* Used automatic cost tracking for 1st time with Propeller Ads, seems to work but of course it won't account for any click loss between traffic source -> tracker, so the cost is usually a bit higher than shown in tracker stats.

Lander performance:


Offer performance:



I did not get to the point of learning to optimize properly but at least there were some other learning experiences along the way

---

Lander CTR is frightfully low! I would DEFINITELY recommend ripping more than 3 landers. Try 5 AND spend the time to make them load faster, function properly, and display properly in all major browsers and devices. I know you said you've already over-optimized the landers - what are the loading speeds?

When doing initial testing, definitely use proven offers - whether they be recommended by AMs as having good CR with more than a couple of other affiliates, or you've seen intel in spy tools that they seem to be exhibiting good trends (received good amount of traffic over more than a couple of days for example - nobody would continue sending vast amounts of traffic to a non-converting offer for too long). You basically have 3 uncertain variables: offers, landers, and traffic. In the beginning you don't have a list of good/bad placements, a proven lander, or a proven offer that you've collected your stats on and can see good performance for, so you need to "sure up" those variables as much as you can.

As for translating your landers, why not just rip landers that were already in Italian? Granted, if you don't know Italian yourself, you wouldn't know how good the copy is anyway, but if you test a few that have promising stats in a spy tool, chances are you'll have at least one good performer.

So - my suggestion would be to get some good offers from your AMs, rip and fix up more landers that have good stats, and repeat the test.



Amy

---

I was able to get the landers loading anywhere between 500ms to 1500ms depending on the test location (using webpagetest). I did minimal checking for how they actually displayed across other devices so I appreciate you making that point!

Edit: Looking back, I think the entire redirect chain was negatively impacted since I tried to use HTTPS for everything (HTTPS tracker -> HTTPS tracking link -> HTTPS lander). This would add extra time cost to each request as I understand now. (More info in this thread: https://stmforum.com/forum/showthrea...her-with-https)


I thought I could be a bit creative and take landers I'd seen running in other geos -- unfortunately the poor result speaks for itself so next time I'll be grabbing some proven landers for sure


Makes sense, I will absolutely do this for the next attempt.


Thanks Amy, I really appreciate your feedback!

---

Interesting!

You know what would be even more interesting? If yourself or someone/anyone would do a split-test on http vs. https and compared the REALLY important metric: Conversion rate.



Amy

---

Home > > Newbie Follow-Alongs