Home > General > Affiliate Marketing Forum

Why split test the exact same offer from different networks? (13)

10-02-2015 07:24 AM #1 1jilliondollars (Member)
Why split test the exact same offer from different networks?

Hey STM

Titles says it all, I don't get it!

I would get it if the payout was much more or less and if some networks had more geo's but even then, affiliate manager's can usually find the geo's they're missing/you want.

I launched my first campaign a couple of weeks ago but right now I got 1 offer I want to promote and 6 networks I am working with that have it too. Having 6-8 banners for each angle and separating restrictions etc adds up to something like 30-40 different campaigns I need to run for the test phase. Whereas I could've launched this within the day if there is no real significance of split testing different networks for the same offers.

So what's the purpose of doing that? Is it the link ? Server location may redirect faster for those geo's?

Thanks in advance,

Aris

10-02-2015 08:29 AM #2 vector (Member)

Here's a good explanation from Charles Ngo of the reason behind split-testing different networks:
The Split-Test Affiliates Forget to do 99% of the Time

Split-testing networks is certainly important but other factors such as offer and angle selection have a much greater impact on earnings. According to Finch in his recent guide (at FinchSells.com), you can add 20-30% to the ROI by finding the best network for the offer. This is significant enough to merit testing but probably a moot point if the angles are not working. IMHO network split-testing is better suited to the optimization phase rather than the launch phase of a campaign when you already have several higher priority variables to test.

10-02-2015 08:52 AM #3 sebastian_r (Member)

Run the split and you'll realise the importance of it immediately.

Use the same landers/banners, just plugin all the offers into your tracking solution. That's 20-30 minutes work max.

If something is not compliant, you can change it later for the winning offer.

I see regularly differences from 20% to 60% in conversion for the exact same offer from the exact same advertiser on different networks. Sometimes even 100% when an offer has reached an network cap or some other bullshit.

10-02-2015 09:24 AM #4 1jilliondollars (Member)

Quote Originally Posted by vector View Post
Here's a good explanation from Charles Ngo of the reason behind split-testing different networks:
The Split-Test Affiliates Forget to do 99% of the Time

Split-testing networks is certainly important but other factors such as offer and angle selection have a much greater impact on earnings. According to Finch in his recent guide (at FinchSells.com), you can add 20-30% to the ROI by finding the best network for the offer. This is significant enough to merit testing but probably a moot point if the angles are not working. IMHO network split-testing is better suited to the optimization phase rather than the launch phase of a campaign when you already have several higher priority variables to test.
That was awesome and included some of the points that were logical. Never knew it could have such an impact though.
His post also brought up some other questions in regards to the networks and their legitimacy. I noticed that 5-6 networks I am apart of have the exact same interface but are apparently completely different companies. I don't know how to identify whether they are good or not.

Thank you

Edit: I just clicked. Charles is talking about the OPTIMIZATION process and not the initial test phase. Makes more sense now I guess. If it will convert well, then it should convert no matter what network you use. You can then scale it to other networks and more traffic sources... This sound good to far?

10-02-2015 09:36 AM #5 1jilliondollars (Member)

Quote Originally Posted by sebastian_r View Post
Run the split and you'll realise the importance of it immediately.

Use the same landers/banners, just plugin all the offers into your tracking solution. That's 20-30 minutes work max.

If something is not compliant, you can change it later for the winning offer.

I see regularly differences from 20% to 60% in conversion for the exact same offer from the exact same advertiser on different networks. Sometimes even 100% when an offer has reached an network cap or some other bullshit.
Thank you for that explanation.

I guess it takes me x10 that time to set everything up because I keep having doubts with what im doing and redoing it over and over again. Each offer has it's own restrictions and rules etc. Not entirely sure how to group the campaigns and sort them out to test it properly.

I guess I'll try to simplify it as much as possible. The only way I can think of doing this is to split test - Mobile / site , mobile/app and wifi/site/, wifi/app for each angle. So, 4 campaigns for each angle for each offer = 12 campaigns for each offer x 6 offers = 72 campaigns It sounded a bit absurd, hence got me thinking if I am doing this right or not.

10-02-2015 12:44 PM #6 cbrughmans (Member)

There's various reasons working with a few networks is a good idea

1. Spread your financial risk

2. Never depend on one external party. What if they lose the campaign? Disapprove all your leads? Go bankrupt? etc
If you have multiple networks set up at the same time you can switch the traffic from one minute to another.

3. Compare stats from one network with another. Compare CTR, CR, eCPM, Margin, etc

4. Most networks give caps to their (new) publishers on most campaigns - WE DONT by the way - so if you have multiple networks you'll get more cap on the same offer

10-02-2015 01:21 PM #7 1jilliondollars (Member)

Quote Originally Posted by cbrughmans View Post
There's various reasons working with a few networks is a good idea

1. Spread your financial risk

2. Never depend on one external party. What if they lose the campaign? Disapprove all your leads? Go bankrupt? etc
If you have multiple networks set up at the same time you can switch the traffic from one minute to another.

3. Compare stats from one network with another. Compare CTR, CR, eCPM, Margin, etc

4. Most networks give caps to their (new) publishers on most campaigns - WE DONT by the way - so if you have multiple networks you'll get more cap on the same offer
Never thought of it that way, thank you for that!

I am a member in Addiliate for about a week now. Will surely work with you soon , thank you

Aris

10-02-2015 01:33 PM #8 cptncrnch (Member)

Quote Originally Posted by 1
If it will convert well, then it should convert no matter what network you use.
Sadly, this is not true. All kinds of things can be going on behind the scenes to alter your CVR from network to network drastically. Not sure why you need 72 campaigns to test this. Can you not rotate offers in your tracker?

10-02-2015 01:45 PM #9 thebrent (Member)
Why split test the exact same offer from different networks?

Certain networks get their leads "throttled" far more by certain advertisers.

Example:

Network A has a long standing relationship with offer X. The owner of offer X decides for network A that he's not going to shave leads and he's going to let new affiliate id's track properly.

Network B has a long history of sending poor quality to offer X, so he works with sales to "throttle" leads so it's profitable for him. New affiliate subs may or may not track, depending on his setup.

We're dealing with computers and postback pixels. They rarely make mistakes, unless there is a tech glitch. So the only thing that explains an offer converting better on a particular network is dependent on who the advertiser is shaving.

10-02-2015 02:07 PM #10 1jilliondollars (Member)

Quote Originally Posted by cptncrnch View Post
Sadly, this is not true. All kinds of things can be going on behind the scenes to alter your CVR from network to network drastically. Not sure why you need 72 campaigns to test this. Can you not rotate offers in your tracker?
Hey and thanks.

Yes I use Voluum and i can rotate but there's different geo's as well and not all networks include the same geo's. I made a mistake on the calculator and got 72

Should be much less:

The tracking links will look like this:
Traffic source - country - Angle 1 - Mobile/App
Traffic source - country - angle 1 - mobile /site
Traffic source - country - angle 1 - wifi/site
Traffic source - country - angle 1 - wifi/app

Traffic source - country - Angle 2 - Mobile/App
Traffic source - country - angle 2 - mobile /site
Traffic source - country - angle 2 - wifi/site
Traffic source - country - angle 2 - wifi/app

Traffic source - country - Angle 3 - Mobile/App
Traffic source - country - angle 3 - mobile /site
Traffic source - country - angle 3 - wifi/site
Traffic source - country - angle 3 - wifi/app


That's for 1 GEO, 1 network.

10-02-2015 03:19 PM #11 caurmen (Administrator)

Very short version: because it makes you more money.

In the past I've tested an identical offer across three networks, and have routinely seen statistically significant results like this:

Network 1 CVR: 6%
Network 2 CVR: 4%
Network 3 CVR: 12%.

The difference between an offer converting at 12% and one converting at 4%, in the same campaign, is... quite large.

10-02-2015 04:06 PM #12 sebastian_r (Member)

Quote Originally Posted by caurmen View Post
The difference between an offer converting at 12% and one converting at 4%, in the same campaign, is... quite large.
If competition runs the 12%er, the 6%er will see no green land. simply not able to spend competitive CPM. It makes or breaks the campaign.

10-02-2015 05:20 PM #13 1jilliondollars (Member)

Got it! Thank you all for your replies.

Crystal clear!

Aris

Home > General > Affiliate Marketing Forum