28 february 2020 1 comments

Make India Hard Again. Part 2: The First Test Results

Hopefully, you have already had the opportunity to read the first part of our special project on driving traffic to India. Today, we are going to discuss how we picked up an offer and creatives and launched our first ad campaigns. We will also touch upon our first results. 

Ad campaigns on trafficstarts.com

We created several ad campaigns based on a source type (PRIME, RON) and device (mobile and desktop traffic):

  • web-MaxX-COD-IN-PRIME
  • mob-MaxX-COD-IN-PRIME
  • web-MaxX-COD-IN-RON
  • mob-MaxX-COD-IN-RON

We didn’t set up advanced targeting options for tests. When the campaigns were already running, we made the following adjustments:

  • we added a blacklist of IPs;
  • we stopped RON and desktop campaigns because costs per lead were too high, so, only Android campaigns were left;
  • we ruled out such categories as Gay, Dating, Shemale.

We decided to place ads only on two websites, namely xhamster.com and sexvid.xxx. Then, only xhamster.com was left (placement m.xhamster.com 2812)

Also, we tested out our hypothesis on dividing campaigns by sections (4 blocks) and time intervals (2 slots). Eventually, we had to abandon it, as the average CR turned out to be lower than that of a standard campaign. Another disadvantage of this option is that it is difficult to update the costs if you work with multiple campaigns and have to do everything manually. All our campaigns were CPM-based, and so we had to update the costs manually for each creative to get accurate data on the performance of each banner.

Choosing an offer

We picked up the following offers:

  • MaxX (17164)
    • pre-landers: 9387 10977 11006 11018
    • landings: 5528 10482 10809
  • Shilajit (18466)
    • pre-landers: 10932
    • landings: 10899
  • MusliPro (18495)
    • pre-landers: 11254
    • landings: 11248

There is no need to save and then upload pre-landers and landing pages to your hosting, as the affiliate network allows you to create separate flows: first, you choose an offer, and then you pick up your pre-lander and landing. After that, you’ll get pre-lander and landing page links. You can add them either to a tracker as separate offers or to your ad campaign on an ad network.

I added the flows to Binom Tracker as separate offers:

It’s up to you to decide whether you are going to use a tracker for campaign optimization or not, but remember that:

  • you can lose some traffic when switching between a traffic source and a tracker (our tracker is hosted on DigitalOcean, and there was no noticeable traffic loss);
  • you should find out where it is more convenient for you to analyze your campaign stats;
  • a tracker allows you to filter out unwanted traffic and add it to a blacklist.

The tests revealed that the MaxX offer performed the best:

Creatives

We used various approaches, but two of them proved to be the most effective:

  • an Indian celebrity or black person with a big penis;
  • a close up of a dick.

The CTR of creatives varied from 0.3% to 1.3%. But at the end of the day, you should focus on the average cost per lead for each creative. Also, bear in mind that an ad network and a tracker display all the conversions you have generated. However, a considerable part of these gets sorted out by an advertiser, especially when it comes to GEO India. We will talk about this in more detail below.

First results

As for now, the stats are as follows:

    • Total leads: 2014
    • Pending leads: 1047
    • Approved leads: 121 (9.29%)
    • Rejected leads: 135
    • Trash leads: 711
  • ROI: about 0%

I requested data on trash leads from Dr.Cash showing why they were filtered out. Turns out most of them were doubles:

People just left their applications several times in a row, which resulted in a high percentage of trash leads.

By the way, the offer didn’t accept traffic from some GEOs (15 Indian states were stopped). Unfortunately, the network does not allow you to blacklist states, and so I will have to deal with this challenge later on. I will probably collect the IPs of these states and disable them. Or I will send these IPs through a traffic back URL to some dating offer. If you have any ideas, please let me know.

As for the approval rate, it is far from great. Dr.cash claims that the average approval rate is 11.15%, but mine turned out to be lower. I tried time targeting and drove traffic from 11 pm to 3 am and then 25 hours a day, but it didn’t affect the approval rate. So, I’m going to change some IP settings and discuss the results in my next article.

Yes, the average ROI is approximately 0%. Hopefully, I will climb into the green soon. 


The next step is to optimize and scale up my campaigns, as well as to split-test new creatives, offers, and landings. This will be the subject of my next article.

See you soon!


Этот текст скопирован с сайта https://zorbasmedia.com
Author:

Have something to tell about affiliate marketing? Become a ZorbasMedia author!

Featured Articles
Black Friday 2020 on ZorbasMedia You have probably been waiting for this day for the whole year. And finally, Black Friday with its awesome deals and huge discounts is here! We have gathered the best and most exclusive discounts and Black Friday deals so that you get real value for money and maximize your affiliate profits. Please be sure to […]
Placing a Bet on Betting We are launching the next experiment with our partners from Pin-Up. This time we are going to test funnels used for gambling on differenet kind of offers -- Betting. Check out the article to find out how and why we are going to do that.
The Man and the Traffic. Season 2.0 — Promoting nutra, dating, and physical products Hi everyone! This is the Man speaking. A couple of months have passed since I shared the results of my campaigns. Running traffic while being stuck inside with four kids is not an easy job, and so my article comes a little late.
Comments
Savo
29 february, 06:20
Awesome work so far!

Would it be wise to send the unwanted traffic back to a revshare webcam offer instead of dating?
Leave a comment
Ответ на комментарий: ....
This site uses cookies. By closing this message and continuing to browse the site, you authorize their use.