I would love to get a solid process of split testing interests when targeting on Facebook. Now I just go with a large chunk of overlapping interests, but I’m sure there is a better way.
MY CURRENT SPLIT TESTING PROCESS:
I run my own offers which are high pay-out, but since I know how much a lead will make me on average over the course of 3 to 6 months I have my own e-mail submit CPA that I aim for. So an email submit counts as a conversion for me.
My process of split testing ads generally the same:
ROUND 1
1 control image -> 5 completely different headlines (aggressive to tame) all in different ad sets.
SPEND 10X PRICE PER LEAD: See which headline wins, take that one as the control headline
ROUND 2
1 Control headline -> 5 different images
SPEND 10X PRICE PER LEAD: See which image wins, take control headline + control image
ROUND 3
Create 5 variations of the control headline + control image
SPEND 20X PRICE PER LEAD
Then go with the 1 or two winners..
So far this has worked reasonably well for me. The thing is, I would really like to split test interest as well. I think this could get my cost per lead down a lot.
Would really appreciate your thoughts on this!
Nobody? Would really like to know a proper way to do this.
Well, when you want to split-test headlines, images, and interests / audiences, things get expansive very quickly.
Not messy, just large in scale.
At the moment you are prioritising variables - headline then image - which is fine.
In this case, I would simply do the same thing but for each interest segment. More work, more spend, more time required. Result = more data and tighter optimisation to each audience.
Otherwise, if you are looking to filter off demographics/interest clusters first, I would take the 2-3 images/headline combos you know work and test them against gender/age/interest brackets initially to gauge what shows the most potential and then get into deeper headline > image testing from there.
Agreed with Zeno, it's a lot of work.
At the end of the day you could either scale and perhaps have less ROI per campaign, but greater bottom line profit,
Or, you could swim in the details and have very high ROI with less campaigns and less profit.
No one way is right and there's only so many hours in the day.
I like to focus on scale and if a campaign merits dissection, delegate it.
But all in all, your approach sounds fine.
Thanks for the answers champs! Yeah I guess it would be better to make my ads appeal to a broader audience then just change the way I sell in the Autoresponder.