As a Facebook ads consultant, I get to peek into lots of Facebook ad campaigns.
A lot of people are doing lots of things right.
But I see lots of under optimized campaigns also.
One of the biggest mistakes I often see is that people have lots of different audiences that they’re targeting with different ad sets.
Targeting different audiences with different ad sets is great when you’re testing audiences.
But once you’ve already identified your best performing audiences, you’ll want to combine your audiences into a small number of Super Ad Sets.
Here’s why…
Facebook’s algorithm does most of its optimizing at the ad set level.
The more conversions that a particular ad set has, the more data Facebook has to optimize delivery of that ad set to deliver impressions to people who are most likely to convert.
The algorithm also works best with the broadest relevant audiences possible so they have the leeway to find the right people for you.
So here’s what you should do…
After you’ve identified several interests that are profitable for you, create a new ad set that combines all those interests, give it a healthy budget and then pause your previous small ad sets.
Do the same with your Lookalikes. A single ad set with all of your profitable Lookalikes combined.
It’s also a good idea to make sure your ad sets aren’t overlapping audiences.
So if you have an ad set of combined interests, you’ll probably want to exclude your lookalikes.
And of course, always make sure to exclude custom audiences of people who have already converted from all your ad sets.
Give this a try. I think you’ll be happy with the results.
Have questions? Send them my why in the comments 🙂
Awesome article 🙂
I’ve been testing 1% and 3 % lookalike of previous leads. Getting good results on a cold audience for lead ads. 3% excludes the 1% and previous customers of course.
I find it all depends on your funnel and if you’re using video or not too.
Cool! Thanks Aaron 🙂 I have a question regarding this:
“So if you have an ad set of combined interests, you’ll probably want to exclude your lookalikes.”
I create LLA after combined AdSet. If I add exclusion to my AdSet then I will edit this AdSet so optimization may start over again. I would like to avoid that. I don’t like touch good working AdSets.
Instead, is it a good idea to exclude CA of those Combined Interests from LLA?
@ Shane – I always exclude each targeting audience from each ad set. Another good way is to save the audience and check their overlap in the custom audience section 🙂
What do you think about looking at overlap during testing phase (3 days with low $10 daily budget on autobid) ? Does it matter? I mean would you worry about it?
I try to avoid overlap from the beginning to be honest, but that’s a perfect scenario. Go ahead and test for sure, I’m all aboyut test and scale, test and scale 🙂
Hi. I want to ask- is it as effective to use multiple custom audiences at the same time or its better I combine them all into 1 single custom adset??
Hi!
I have a question: is it better to separate Lookalike audience from the Intersests, etc. Audience for real estate?
Thx!
Does that mean that if I want to target 13-18 and 45-60, I should have one Adset that target both ?
Because there’s a gap in the age buckets, you would need to set that up in two separate ad sets. You can’t do that in a single ad set.
Hi Aaron – wonderful article. I’m having a really frustrating time with FB ads. I’ve tested my entire funnel and I know it works but FB just seems so temperamental. eg. I will turn ads of and on and suddenly they perform much better. Then I want to scale and create a new adset using lookalikes and some interest targetting, increase the budget and then suddenly it craps itself. Any thoughts on what I can do to scale more? At present, I’m just using one adset but have found multiple adsets work better but looking for better ways to grow. thanks in advance
Hi aaron! Would like to ask your two cents if which would be more advisable? if we are in the whole of Australia but only certain regions we do free shipping.
1. To separate the campaigns by region? and split the ad sets by different audience targeting
2. To combine all the regions in 1 ad set and different ad sets will have different audience targeting and having 1 campaign.
Aaron,
We just ran a facebook test campaign at $45/day at the ad set level. We ran 3 campaigns with 3 ad sets each. Within those ad sets we had 8 ads per ad set. We spent a lot and got ZERO results.
Here’s my question: Did we run too many ad sets for that size of budget? Would we have better success if we reduced the number of ads sets and increased budget per ad set or should that not matter?
There are a lot of variables at work here so it’s hard to know why it didn’t work. However, I’d focus more about bigger things like your targeting, offer and ad creatives than on the budget allocation. If those other elements are much more impactful.