Home > Affiliate Marketing Forum >

Always Test The Same Offer On Different Networks (23)

11-25-2011 02:56 AM #1 maynzie (Moderator)
Always Test The Same Offer On Different Networks

I am not going to out any of the networks I involved in this study but I was actually pretty shocked when I tested the same offer over 5 networks and the results were a bit scary, cos it could mean that you do have a profitable campaign only the offer is letting you down.

I tested an offer over 4 networks and sent 200 clicks to each. here are the results

Network 1
Payout was 5.50
with 200 clicks I got 11 leads with an epc of 30 cents

Network 2
Payout was 5.15
With 200 clicks I got 14 leads with an epc of 36 cents

Network 3
Payout was 5.25
With 200 clicks I got 7 leads with an epc of 18 cents

Network 4
Payout was 5.05
With 200 clicks I got 27 leads with an epc 68 cents

Well as you can see, network 4 was the heavy hitter and I have no fucking clue to why it converted so so much better then the other networks.

And what was more twisted was this network, the reason why I am not naming networks is always sledged at having bad tracking and that the offers dont convert anywhere as good, and the second lowest epc was a network that's known for having superb tracking??

Well I can tell you it PAYS to TEST on DIFFERENT networks, so I know AM's we all get loyal to them and they are great friends, but if your holding back a campaign because of loyalty remember this is your job and they have 100's of others affs making them money too.

But in the end of the day, TEST DIFFERENT NETWORKS for the same offer as I just turned a slightly losing campaign into a 100% roi campaign.

11-25-2011 06:22 AM #2 agent007 (Member)

Great tip Maynzie. Do you mind mentioning which vertical you used?

I starting to test the same offers with different networks for this very reason. But for now I'll keep it to about three until I can ramp up my volume again in order to get weekly payments asap lol

11-25-2011 09:57 AM #3 lavish (Member)

Although I generally agree with your point... that's a pretty low number of conversions to make a conclusive argument.

11-25-2011 10:24 AM #4 maynzie (Moderator)

Yah I do agree man, but I have done quite a lot of volume to network 4 (over 6000 leads at 13.95% conversion rates) and in the past 2 months I have tested a network 1 and 3 for around 900 clicks each letting it run for 1-2 days and they had a conversion rate of 4-6% looking back at the stats now. I posted the shorter case study that I ran yesterday to demonstrate to the forum the difference between the offers, and for most beginners they cut off a campaign without profits in a single day, but as you can see depending on what network you ran with, would determine whether you ran the campaign or not.

so yeah might not be the best information for the seasoned pro, but it could make or break the motivation for an up and coming marketer

11-28-2011 08:32 PM #5 tijn (Moderator)

great post man.

it would be interesting to see a comparison of load speed / number of redirects for each of the offer links for each network.

11-29-2011 01:30 AM #6 drak (Member)

Great advice, I've also had EPCs practically double just by testing different networks. I've even had some offers that I had a paybump on lose to another network that didn't have a bump yet. Some offers will shave different networks at different times so that could be one cause.

Another thing to remember is to try and split your traffic and test them at the same time, Sunday might be that offers strongest day so if your first network gets Sunday and second network gets Monday then you wont have complete accuracy and could lose out on some money.

11-29-2011 04:11 AM #7 rileypool (Member)

Are you reading the EPCs reported to you by your tracking solution or the EPCs reported to you by the network? I too won't name networks, but some networks manipulate their inbound clicks, so if you're looking at the EPC on the network, it'll artificially be higher than it really is.

11-29-2011 06:30 AM #8 tijn (Moderator)

Quote Originally Posted by drak View Post
Another thing to remember is to try and split your traffic and test them at the same time, Sunday might be that offers strongest day so if your first network gets Sunday and second network gets Monday then you wont have complete accuracy and could lose out on some money.
I have found this with email submits a lot where you have two exact same offers but on different networks and they seem to flip positions. One does better in the morning where the other performs better at night. Or one performs better at the start of the week and the other at the end/weekend.

11-29-2011 02:11 PM #9 Smaxor (Veteran Member)

All this is a great point. There are so many factors in things. Always remember that you're in with a bunch of other affiliates on an offer. If the other affiliates traffic sucks then there's a chance that maybe the advertiser is shaving to compensate for it and you're being penalized for something else someone else is fucking up. Also as someone mentioned, network load times are important that's a HUGE piece of the puzzle.

One thing I can say is split between networks then if you have a network you particularly like, and they're not the winning network, just communicate with them the situation and see if they can get to the bottom of it.

P.S. one other thing when you do these comparisons make sure you're not using the click count from the network. There's a handful of networks around that shave clicks to make epc's look better when in reality they're the same. Make sure you're using your clicks from your own tracking to calculate EPC. But yes always split test

11-29-2011 02:39 PM #10 thekaine (Member)

thing is, why should you bother going that road?

When you do not test it you will never find out (so you never complain and just stop the campaign).

When you DO find out about it by running it on different networks and get statistical relevant data and see that one is better, i would not go back and ask the other network(s) why it is like that. For me there is only the money difference and it might be true thats not the Networks fault in general but the Advertisers that are running the offer but then again ... for me its still just the monney difference.

Would be super awesome when networks would tell you right away "oh okay we drop like 50% of the conversions atm because our Affiliates are doing shit" then you can show quality traffic and get different payout or w/e. But as it is atm neither the networks have full clue nor does those that are running the offer ...


Makes it quite a bad thing :/

11-29-2011 05:20 PM #11 polarbacon (Moderator)

Quote Originally Posted by thekaine View Post
thing is, why should you bother going that road?

When you do not test it you will never find out (so you never complain and just stop the campaign).

When you DO find out about it by running it on different networks and get statistical relevant data and see that one is better, i would not go back and ask the other network(s) why it is like that. For me there is only the money difference and it might be true thats not the Networks fault in general but the Advertisers that are running the offer but then again ... for me its still just the monney difference.

Would be super awesome when networks would tell you right away "oh okay we drop like 50% of the conversions atm because our Affiliates are doing shit" then you can show quality traffic and get different payout or w/e. But as it is atm neither the networks have full clue nor does those that are running the offer ...


Makes it quite a bad thing :/

hmmm....imo your missing out...at the very least if gives you a chance to beat up a network...and what I have found out sometimes that there have been pixel issues (very very common).....you never know....its not always about the money.....

some networks are pretty cool....and really work hard to make shit work for you.....
I try to always give people a second chance.....

just my 2 cents

11-29-2011 06:26 PM #12 thekaine (Member)

well i would not leave the network because of that and would definatly run there stuff again. Its not like i would "blacklist" networks after stuff like that just talking strictly on a offer base

But yeah the pixel thingy is a thing to consider even though when i think about it the network should have that in mind all the time and test it or something (working for a big merchant in my country and we run pixels aswell and i have to check them every day and make sure they are working etc ...)

11-29-2011 06:34 PM #13 illya (Member)

Quote Originally Posted by thekaine View Post
thing is, why should you bother going that road?

When you do not test it you will never find out (so you never complain and just stop the campaign).

When you DO find out about it by running it on different networks and get statistical relevant data and see that one is better, i would not go back and ask the other network(s) why it is like that. For me there is only the money difference and it might be true thats not the Networks fault in general but the Advertisers that are running the offer but then again ... for me its still just the monney difference.

Would be super awesome when networks would tell you right away "oh okay we drop like 50% of the conversions atm because our Affiliates are doing shit" then you can show quality traffic and get different payout or w/e. But as it is atm neither the networks have full clue nor does those that are running the offer ...


Makes it quite a bad thing :/
Discussion can have positive outcomes as well. You dont have to accuse anyone of anything.

I started a conversation with 2 networks about quality because of my concern about the conversion rate. The payout kept bouncing and conversion rates were inconsistent. The networks claimed it was due to overall offer quality (even though MY setup was consistent). They didnt admit the conversion rate was affected by quality but this probe opened the door to...

The networks asked the advertiser to do a qc check on just MY campaigns, then I got a 'private' offering from both networks with my own 'special' payout.

BTW. Coincidentally, my conversion rate improved.

This was possible because I was pushing a lot of volume and had a good rep with both networks.

This is testimony to Alex's (angryrussian) advice that volume has more positive effects on your bottom line, and can trump short term ROI.

11-29-2011 09:07 PM #14 DarkoM (Member)

Nice Thread!!!

I never do that....test just the networks. Idiot!

I love this forum. I learn every day here

12-02-2011 08:32 PM #15 mphilips (AMC Alumnus)

Quote Originally Posted by Smaxor View Post
All this is a great point. There are so many factors in things. Always remember that you're in with a bunch of other affiliates on an offer. If the other affiliates traffic sucks then there's a chance that maybe the advertiser is shaving to compensate for it and you're being penalized for something else someone else is fucking up. Also as someone mentioned, network load times are important that's a HUGE piece of the puzzle.

One thing I can say is split between networks then if you have a network you particularly like, and they're not the winning network, just communicate with them the situation and see if they can get to the bottom of it.
Smaxor's got it right and that's exactly what i do. I have a few favored networks but I always split test with as many as I can. Any time you run a new campaign you should do this to rule out the possibility that (if the campaign isn't work that well) it is the network causing it to fail and not your work.

10-01-2012 08:51 AM #16 maynzie (Moderator)

I've been helping an affiliate out recently and he said he couldn't get the campaign to work, so I told him to split across a few networks with the same offer, so he tested out 3 networks and 1 of them converted triple the amount of the lowest performer and 1.5x the second performer, this took a campaign from the negatives into the positives.

Can't stress enough how important this is to make sure you do split across networks, especially if you're running volume.

+ In the original post I forgot to mention and it was stated in this thread, always go by total revenue over epc, some networks don't report double clicks or even yet shave clicks, so epc's aren't always truthful. Better comparing the revenues

10-01-2012 09:21 AM #17 qhead (Member)

Quote Originally Posted by maynzie View Post
I've been helping an affiliate out recently and he said he couldn't get the campaign to work, so I told him to split across a few networks with the same offer, so he tested out 3 networks and 1 of them converted triple the amount of the lowest performer and 1.5x the second performer, this took a campaign from the negatives into the positives.

Can't stress enough how important this is to make sure you do split across networks, especially if you're running volume.

+ In the original post I forgot to mention and it was stated in this thread, always go by total revenue over epc, some networks don't report double clicks or even yet shave clicks, so epc's aren't always truthful. Better comparing the revenues
I have been having exactly same experience with the current campaign I'm doing. Another thing is that different networks seem to convert differently for different age groups (in dating). I have my own theory why but not enough concrete data so I'm not going to speculate

10-01-2012 01:35 PM #18 dr_ngo ()

Nice tip, can definitely make or break campaigns. If the network has lower EPC's, it might not be them scrubbing you.

Sometimes the advertiser will scrub the network as a whole to make up for bad leads or whatever.

10-01-2012 02:29 PM #19 Smaxor (Veteran Member)

Just went back and read this again. One thing you want to make sure of is that you're getting enough data to be relevant. 200 clicks is pretty low.

You can use this calculator to make sure that your data is significant enough to matter.

http://www.splittester.com/index.php

10-01-2012 02:48 PM #20 thmp (Member)

But that's when you are driving a lot of volume, right?
I mean, if it is only a few clicks a day, like 20, it's pointless to split test traffic between networks.. to gain statistically significant data, you'd need a lot of clicks.

10-01-2012 03:22 PM #21 qhead (Member)

Quote Originally Posted by thmp View Post
But that's when you are driving a lot of volume, right?
I mean, if it is only a few clicks a day, like 20, it's pointless to split test traffic between networks.. to gain statistically significant data, you'd need a lot of clicks.
Statistical significance can be achieve with quite low amount of data. There's no clear cut rule when you have enough data, the math behind statistical confidence on outcome depends on how different results the variants are pulling.

10-01-2012 07:43 PM #22 thmp (Member)

Quote Originally Posted by qhead View Post
Statistical significance can be achieve with quite low amount of data. There's no clear cut rule when you have enough data, the math behind statistical confidence on outcome depends on how different results the variants are pulling.
Ok, thanks man..

02-16-2014 05:55 AM #23 iAmAttila (Veteran Member)

I know this is bumping an old post, but the day I did what OP suggested (last december), is the day I started 'getting it' and making profit.

I was basically drinking the kool aid of some self claimed affiliate marketing magnate; one he feeds to his members on his paid forum and believed all of what he said - that his network is the best out there, and no other network comes close! It was a real LOL, and OMG moment when I split tested same offer on that network with CD & MUNDO and both kicked their ass cvr wise.

2 Lessons learned

1) Don't trust what people say, just because...
2) Always split test networks every single time

Home > Affiliate Marketing Forum >