A couple posts back I wrote about scaling ppv campaigns. One of the ideas was to split test offer across different CPA networks. I wanted to expand on that idea and show an example from one of my campaigns where this made a huge difference.
Affiliates often get blindly loyal to one affiliate network, which can actually really put a damper on their progress. This is pretty normal for most affiliates but
Some of the reasons affiliates stick with one network
– You become friends with your AM and feel loyal to them.
– You go to a party thrown by them at ASW,or get taken out to dinner.
– You just run there because ‘everybody' says it's such a great network, or it's the one you hear about all the time.
Those aren't necessarily bad reasons, but business is business and ultimately it's not going to help you build your business if you don't try an offer at another network because you feel like you're cheating on your AM or something.
Why would an offer convert differently at another network?
I get asked this question a lot, and I can totally understand why. It doesn't seem like an offer should perform any different on another network. I mean, it's the same offer right? Here are a couple reasons that offers can convert better on one network vs. another.
1. Redirect speeds
Redirect speeds are basically how fast the traffic goes from your link, through the CPA networks tracking system to the offer. Redirect speed can be affected by things like hosting, type of tracking platform the network uses, etc.
2. Brokered networks
A ‘brokered' network is one that runs it's offers from another network. Network ‘A' might have a diet offer, and network ‘B' basically brokers that offer and it's publishers run it. If you think about that in terms of the redirect speeds, you can see a potential problem.
When you run an offer from a network that is brokered from another, it has to go through network A, to Network B then to the offer. This can really slow things down and cause the offer page to take longer to load. This is also an opportunity for cookies or tracking to drop off which can result in lower conversions.
Did you also know that there are brokered, brokered networks? These are networks that broker their offers from another network who is already brokering offers.
3. Scrubbing
Ah, the dreaded scrubbing. I want to share something a little different than you may have heard about scrubbing though. What some affiliates don't realize is that the advertiser can scrub the CPA network. Why does this happen? Usually because the network is sending bad quality. Some advertisers have a formula for this like ‘if a CPA network has promoted for x number of days with a conversion rate below y, the scrub rate increases to z. The advertisers aren't necessarily out to screw anyone, sometimes they need to set a scrub rate to remain profitable.
So basically, if a CPA network is sending crap traffic to an offer, there is a chance they could be getting scrubbed. How do you know? Well, you don't really which is why I recommend split testing your offer across networks.
I'll share the results of a test I did recently
This was the exact same offer and offer page on 3 different networks. The traffic source, keywords/targets, etc. did not change during this test. This was set up in CPV Lab to rotate these 3 networks evenly.
Network A – Conversion rate 4.35
Network B – Conversion rate 6.49%
Network C – Conversion rate 3.70%
You can see the conversion rates varied a lot. I see this all the time when I split test offers like this. When I see something like this I'm pretty direct with the network I don't run it on, and tell them exactly why. No reason to hide, or feel guilty…this is just about the bottom line. Some really honest networks have even suggested I run an offer somewhere else if they know a similar offer is doing better.
Those are my ideas about why it's so important to split test networks. I'm curious to hear if you guys have seen similar, or if you haven't split tested this yet?
Yep, I have seen the same exact thing. I didn’t know that advertisers scrub networks different though..interesting.
So how can you tell which networks or offers are brokered or not?
Redirect speeds are a huge one for me. As soon as I got a decent server that was properly tuned, and optimized my pages for load time my ROI increased. That doesn’t help the networks conversion rate of course, but every bit helps.
Yeah, all kinds of stuff going on behind the scenes!
I just edited the post to include this but you can use a Firefox extension called Live HTTP Headers to show you all the redirects.
I totally agree…increased speed is a good thing anywhere you can get it.
So is there a way to research this BEFORE you begin a campaign.
Or
Will you just have to stick to the hard way, split testing.
Hey David. Are you finding that your results are specific to an offer? Or do you find that if Network C is performing bad with one offer, that it’s consistently bad across the board with all their offers.
Well, ultimately you will never know until you test. You can use things like Live HTTP headers to check how many redirects an affiliate offer has to go through. Once you do this a few times though you’ll find that certain networks consistently perform better and that helps eliminate a lot of the guesswork.
Yeah sometimes it’s specific to the offer. For example one network could be getting scrubbed by the advertiser because their affiliates are sending bad leads, but the rest of their offers are ok. Having said that..I do find that certain networks just generally convert better than others overall.
David great tips.
I’ve come across the same thing in PPV:Lab recently:
Network A – Conversion rate 15.38%
Network B – Conversion rate 6.38%
Network C – Conversion rate 32.05%
I actually split test the traffic for one offer across 4 networks. The last network didn’t convert at all.
Just because “Scrubbing” happens does not always mean you have to deal with it. Some CRM’s offer services to help prevent this.
How can you prevent scrubbing that happens on the advertisers side?
Page load speed really helps. I just hired a webmaster who is also an expert site admin and he reconfigured my site. After two weeks my site increased in traffic and it on sales. But that’s already a good improvement.
Great post man.