Facebook Targeting Optimization Question

[johnp]

I have been running Facebook ads to an offer for cold traffic. My audience is about 450k, to be honest, I kind of screwed up my initial targeting when I first started, but I don't want to pause everything just yet.

For some reason, I'm really getting hung up on what to do with all of the data that I'm getting back with interests and demographics that are producing the best results. So here's my question.

Let's say that I find 3 interests that are converting great. Which of the following should I do?

Option 1) Take the 3 interests out of that adset and combine them into a different adset in the same campaign?

Option 2) Take the 3 interests out of that adset and build a new campaign?

Option 3) Start changing the actual live adset by fine tuning the targeting. I would think that this could be bad though because the ad will have to get approved again.

Option 4) Not do anything and let the Facebook pixel figure everything out for me?

Option 5) None of that, something else?

Thanks!

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[Chris McCarron]

Leave the adset alone, then create multiple variations with different images and headlines.

I'm using Noah Kagan's technique and here are the results I'm getting:

The crazy thing is that your present audience might be really profitable. As you can see above, the only difference between me paying £3.13 per click and £0.33 is the image. That blows my mind.

So test what you have and if an ad isn't profitable... kill it.

 

[johnp]

Thanks for the response. I'll watch through his video. I love Noah's content but I didn't think that Noah was a traffic guy.

To be honest, some of your results don't seem that great. But I'm not sure what you are promoting, so maybe it doesn't matter.

I have heard from other people not to do what you are doing because Facebook will sometimes pick the best ad too fast. Personally, I don't split test my ad copy like that when running ads on a low budget. There are so many different variations, you would have to throw a lot of money at the wall to get statistical significance. I like to test one or two ads at a time, look closely at all of the results, then adjust.

I'm not really interested in split testing a ton of different ads right now. This thread is more about how to optimize the demographics/what to do with interests that you find that are doing better.

 

[GuestUser450]

I have been running Facebook ads to an offer for cold traffic...

Is there a sequence (lander, optin, offer, retarget) or is it straight to an offer?

Are you negative, break even or profitable?

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

[johnp]

Is there a sequence (lander, optin, offer, retarget) or is it straight to an offer?

Straight to just free + shipping for now with backend sales, follow-up through email

 

[johnp]

This article sort of summarizes what I have been doing. I just found it. It seems like I'm doing the right stuff. Although, I won't always limit my interest size to just one interest.

A Crash Course on A/B Testing Facebook Ad Campaigns

Assuming the advice is good and accurate, I think my question is answered.

 

[GuestUser450]

I think my question is answered.

Good luck.

Only thing I'd add is that if you have a good-sized list of people who've converted is do lookalike campaigns - parse them by geo and split-test them against each other. Running multiple ads in a single ad set can handicap optimization like you said - Facebook will decide what it wants to run for you.

 

Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.