different traffic sources, different results (7)
Hey,
just wondering what your thoughts are on this "phenomena".
I have a couple of landers that I tested on Decisive which let's say performed B>C>A.
I tested the same landers with the same banners on another traffic source, same targeting but the results were dramatically different. Here, A performed way better than C&B.
Now, I understand that different traffic sources may have different sites/apps but they are still humans lol so what gives the huge difference in response?
What do you think about this?
Haha, you have discovered Arrow's Impossibility Theorem for affiliate marketing! You deserve a Nobel Prize!
The demographics could be completely different (age/gender/interests/etc)
The placements could be completely different
The position that traffic source fills at could be completely different (Traffic source 1 might get 80% of the impressions, while traffic source 2 gets only 20%)
I don't run too much mobile, so can't really give specifics, but that's what stands out to me
that's also my guess that because of the different apps/sites the demo- and psychographics are different and thus the response to the different marketing messages.
What does that mean for us affiliate marketers?
Do we have to test our packages individually for each traffic source or is this just something that happens sometimes?
You guys don't seem to have made the same experience. I'm assuming you do run traffic from different sources lol
Originally Posted by cmdeal
affiliate marketing = test test leave no testing stone unturned
There may be a single advertising message for your offer that would persuade a 22-year-old female entrepreneur in Silicon Valley who's checking her stocks in a two-minute boardroom break, a broke 80-year-old man farming sheep in the outskirts of Turkmenistan who just got given a phone by his grandson, and a bored 40-something Japanese housewife playing games to all install your app, but it'd be the exception rather than the rule.
Generally, angles won't work on "all humanity". Even when we're talking about a successful broad angle, it's successful within, say, a particular group of Android phone users within the US, which is actually a pretty narrow demographic when you're considering all the possibilities.
Here's one specific example: Grindr. Grindr's one of the most blocked placements, despite the fact that its traffic is actually almost all human. Why? Because there's a very specific demographic and a very specific use intent there, and the messages that work on broad campaigns generally don't work on Grindr - you need to tailor your message to the demo and usage intent.
Thanks for the mental travel around the world 
I understand that you can only skim threads for the most part judging by the astonishing amount of value you put out here. I mentioned that I used the same targeting & the same banners and we already figured out that the demographic and user intent must be different because of different publishers in the networks even though we speak to "the same people".
The interesting implication is when it comes down to "scaling" your campaign. I'm on a tiny budget so I had only very limited experience so far, but from what I've seen it seems like testing your creatives is heavily dependent on the traffic source you're using. A campaign that bombed on network A may be profitable on network B (both DSPs).
Are we really optimizing for the traffic source? What are your thoughts on this?
Personally, I will stay away from "black box" networks and dig a little beneath the surface from now on. Instead of looking at networks or geos in general I try and take a close look at the profitable placements and try to find them (and similar) on other networks.
Home > General > Affiliate Marketing Forum