The Man and the traffic: split tests and Smartlinks

27 August 2019
37
0
Reading: 8 min

The Man and the traffic: split tests and Smartlinks

Hi everyone! This is the editorial team speaking. You’re reading the tenth anniversary volume of our experiment “The Man and the Traffic”. It has been two weeks since the Man started driving traffic to dating offers. We have been following his ups and downs, one day his campaigns were generating profit, another day they plunged into the red. However, the Man doesn’t give up. He has set himself a goal — become a top affiliate — and he has been consistently pursuing it ever since.


Last week, you voted for a new vertical the Man should start working with. Here are the results of the vote:

The Man and the traffic: split tests and Smartlinks

Mikhail Funk, the Man’s mentor, recommended split testing different offers, while the affiliate network manager suggested testing a Smartlink once again. In this volume, the Man discusses how it all went.

Introduction

Last week, I planned to split test various offers targeted at the same GEO and take three or four offers for each region. At that, I decided to launch the tests in the evening to get the maximum traffic volumes and then test a Smartlink once again. The end goal was to make the campaigns I had already launched profitable.

It turns out it’s not that easy to ensure a positive return in affiliate marketing. I’m still not there yet, but my ROI gradually continues climbing out of the red. I have set myself a goal and I’m going to pursue it no matter what: if others can make it, I can make it too. My main concern now is the lack of time. By the way, my wife has come up with a great approach to waking me up: she just tells me at 4 hours in the morning, “Wake up and go earn some money!” And it actually works. But…

The Man and the traffic: split tests and Smartlinks

How I make creatives

In this volume, I would like to describe the process of making creatives in detail. Probably some of my ideas and approaches will come in handy for you, and I also hope that you will give me some advice on simplifying and streamlining this process.

Currently, I use spy service Anstrex to make creatives. I have already elaborated upon its features before. To start analyzing what your competitors are running, you should select a language or a GEO. Having studied push notifications, as well as texts and images used by your competitors, you can examine each creative more thoroughly. Then, I copy some of the texts and their respective translations into Excel.

If you open your competitors’ push notifications in the Creatives tab, you will be able to see what GEOs they targeted:

The Man and the traffic: split tests and Smartlinks

Here, you can also examine the full list of GEOs and some sample approaches in more detail.

When I’m making creatives, I follow these steps:

  1. Analyzing what my competitors are running and organizing all the data in Anstrex. I divided all dating creatives into three categories: direct appeal, a message in a messenger chat, explicit reference to dating.
  2. Studying a particular GEO. When making a selection of creatives, I consult the following articles: the most popular messenger apps by GEO and adult searches by country.
  3. Searching for and creating images. I do searches in Google Images using such keywords as sexy girls, sexy girls boob, Asia girls and so on, as well as messenger icons and some other stuff. You can edit images in Photoshop or any other graphics editor.
  4. Creating texts. I use two approaches: I either translate set phrases into a GEO’s language through Google Translate or use some texts from pre-landers or websites shown in the search results. Let me explain: you enter something like “joder gratis con mujeres” into the Google search bar and get the following search results:

The Man and the traffic: split tests and Smartlinks

Below is one of the doorway sites where you can also find some ready-made texts:

The Man and the traffic: split tests and Smartlinks

I would appreciate it if you would share some of your approaches to making creatives.

Test results

To run split tests, I set up similar campaigns on traffic sources MegaPush and PropellerAds taking into account the features of each source.

Smartlinks:

I launched test campaigns targeted at Malaysia, Taiwan, and Thailand. As adult traffic is prohibited in both these sources, the traffic was driven to offer 57990 [WEB+MOB] MAINSTREAM DATING SMARTLINK /GLOBAL.

Malaysia generated the largest number of leads (30) and Taiwan was a runner-up (27 leads). All the campaigns brought a negative ROI (ROI -80%), as the payouts for this offer were low.

The Man and the traffic: split tests and Smartlinks

Almost all Megapu.sh campaigns exceeded the spending limit of $2 I had set. As a result, I spent up to $8-10 on some test campaigns.

Conclusions: this particular offer is good only for cheap traffic (such as traffic back, for instance).

Split testing offers:

Mexico: I tested three offers:

  • 44348 [MOB] Setravieso 153728 /MX 25+ SOI
  • 64177 [WEB+MOB] MadurasMexico 188517 /MX CPS
  • 69341 [MOB] DivierteteLigando 198151 /MX 25+ SOI

The 69341 offer had the highest conversion rate taking into account all the campaigns. It was exactly this offer that my ClickDealer manager recommended driving traffic to.

Malaysia:

  • 71759 [MOB] IGetNaughty 202340 /MY SOI 25+
  • 58182 [MOB] BeNaughty 186214 /MY SOI 25+
  • 58182 [MOB] BeNaughty 178566 /MY SOI 25+

71759 performed the best. The landings I used are given below:

The Man and the traffic: split tests and Smartlinks

Germany:

  • 53770 [MOB] Amateurcommunity 170411 /DE SOI 1,7 USD
  • 52887 [MOB] Xpartner 168759 /DE SOI 2,3 USD
  • 68562 [MOB] FlirtHits 203150 /DE SOI 1,8 USD
  • 64672 [MOB] CheekyLovers 189495 /DE SOI 25+ 3,1 USD

53770 had the highest conversion rate. The landings I used are given below:

The Man and the traffic: split tests and Smartlinks

Conclusions: Split testing confirmed that offers and landings that were recommended by affiliate network managers had a higher conversion rate. Trust your managers and keep in touch with them, as it will help you save time and money [we talked to an account manager, and he is of the same opinion – Ed.].

Just a brief reminder: I work with ClickDealer, an affiliate network with lots of offers targeted at various GEOs.

The Man and the traffic: split tests and Smartlinks

There are two types of accounts under this affiliate network: “Regular” and “Smartlink Only” Holders of the latter account can apply for an account type change and submit additional information to get access to the full portfolio of offers.

When setting up campaigns on PropellerAds, I always use Target CPA, a tool for automatic optimization of push campaigns:

The Man and the traffic: split tests and Smartlinks

This functionality helps you avoid spending too much time on optimization and, more importantly, it cuts off not just zones but poorly converting slices (zone+GEO+device).

Having split tested the offers, I created and launched ad campaigns targeted at HK, TW, and TH. I’m also planning to scale up the campaigns later.

The Man and the traffic: split tests and Smartlinks

Mentors’ advice

Mikhail Funk

A push traffic expert

The Man and the traffic: split tests and Smartlinks

1. You don’t test enough. The fewer offers you test, the less likely you’ll find a profitable campaign. Offers always presuppose different target actions. As for dating, you can use various pre-landers and strive to make users subscribe to push notifications, fill in a lengthy registration form or verify their email address. To choose the best performing stuff, you should run extensive tests.

2. You should use simpler creatives. As for sweepstakes, it can be something like “You have won a super prize”, “There’s a parcel for you” + the logo of a local post office. And you should also split-test as many sweepstakes targeted at a particular GEO as possible.

Roman S.

An affiliate marketing expert at PropellerAds

The Man and the traffic: split tests and Smartlinks

Hi!

The way I see it, you are evidently making progress. However, there’s always room for improvement.

I would recommend searching for new approaches to making creatives and localizing them for a particular GEO. Add the flags of the countries you’re working with to your creatives, test different girl types. And pay more attention to images: 70% of success when it comes to creatives is the image, and the remaining 30% is left for the text. Spy tools will help you with that. Unfortunately, this time you haven’t shown any creatives, and so it’s difficult for me to comment on the results.

You have started using survey landing pages, just as I have advised you to do. Don’t forget to check whether it’s displayed correctly on mobile devices and how long it takes to load a page — a landing should be opened in less than 2 seconds.

I have noticed that you choose 25+ offers, thus cutting off a huge share of young people. Try driving traffic to 18+ offers, which will allow you to cover a wider audience.

As for verticals, you were right in paying attention to sweepstakes. I recommend driving traffic to sweepstake offers that accept leads. Choose something that will appeal to a wider audience: iPhone giveaways, Amazon cards and so on. Such offers will allow you to figure out the ins and outs of this vertical and test new approaches. Try to find an offer that will appeal to both men and women. As regards creatives, you can use a nice picture of a product. Bear in mind that you should use a vivid high contrast image, as a push notification must immediately attract users’ attention.

Plans for next week

What I’m planning to do is:

  1. Analyze what my competitors are running, make creatives and start testing sweepstakes.
  2. Launch new dating campaigns and start driving traffic to new GEOs after analyzing what GEOs my competitors’ are running traffic to.

Summary

The resources spent so far:

Weekly money spent (from August 19 to August 25): ad campaigns in push networks: $284

Money earned since the first day of the experiment: ClickDealer balance (the money hasn’t been withdrawn yet): $856

The Man and the traffic: split tests and Smartlinks

Balance: $1172

Time: 113 hours

 

Yours faithfully,

the Man

Have a story to tell about traffic arbitrage?
Become a ZorbasMedia contributor!
Become an author