Testing framework on TikTok Ads

Put On Your Mask
D

[DeC]

Active Member
Hello!

I started advertising on TikTok back in 2021 and it was mess performance wise. Sometimes crazy good returns, but the next day CPAs over the roof. Eventually CPAs skyrocketed to the point that no longer made sense for us to keep spending. Fast forward in 2025/2026, I decided to give it a try again. To my surprise, it looks like nothing has changed here at all. It starts off with great results but creatives fatigue at lighting speed. What used to bring $10-$15 CPAs with ~1000$/day spend now it's $40+ in the span of a couple of weeks.

I know that's how TikTok works, if you don't have a constant flow of new creatives that perform on a daily basis, you're cooked. And now we're cooked because no matter what I try, CPAs are unaffordable.

On Meta everything is smoother. Its own rollercoaster, but at least creatives degrade at a lower pace, leaving you some breathing room. There, I have my testing campaign (targeting Installs), my validating campaign (targeting a custom event) and then my scaling campaign (targeting same custom event with higher budget) with good success so far.

So I'm wondering, how are you guys testing TikTok Ads? I'm mostly interested in the structure and framework you guys follow. I've tried to replicate what works for me on Meta Ads, but nope, creatives burn at a higher pace than I'm able to test them.
 
You have to rotate assets on TT almost every day/two days if you're spending the whole budget with no caps.

I normally have an event campaign (never installs) where I run a testing ad group and a BAU ad group at the same time and every day I bring rotates from testing to BAU if performance is aligned.

Starting the campaign with a cap can be also a good idea if your CPAs tank quickly, you may spend a bit less per day but that should delay the fatigue.

Also, don't assume that Meta videos are gonna work on TikTok. Videos with classic transitions and so on don't work there. You have to go with videos that look organic (UGC)
 
Thanks for the reply, specially the cost cap that I'm going to setup right now. I prefer to spend less but with numbers that make sense. I'm a bit tired of burning money.

Under the same campaign, two ad groups? One for testing and the other for scaling? Do you have any rules as in min. spend or min. impressions? On Meta I promote ads that at least have 10K impressions each with metrics that meet my criteria, otherwise directly to the bin.

As for creative velocity, how many ads do you normally run on your BAU? Any differentiation in creative angle or everything to the same ad group?

As for creative tests as well, how many do you normally run on a weekly basis with a similar budget?

Thanks!
 
I put 2 ad groups only i have enough budget to generate around 10 events per day - otherwise i do the rotation within the same BAU ad group. As you said, even winning creatives get fatigue pretty quickly on TT so it's a matter of time that you simply have to exchange that winner for a completely new video.

I rotate every time that my asset get more than $100 of spend (although this minimum depends on how much is your CPA in average). What I do is to take the winning asset performance as a reference and if after spending enough money to generate 2-5 conversions, performance is not close to the reference, I just pause the asset and keep rotating.

The higher budget, the higher velocity rotating because more ads will get the minimum
 
@DeC
I feel you. Have exact the same experience. Meta is so smoth and TT screams right after I clicked on "publish" for new creatives.😭
My graphics department can't keep up with it all.

Sometimes I feel that TikTok doesn't always tell us the truth. For example, I noticed that I had two campaigns with the same number of creatives but different daily budgets. One was set at 250€ and the other at 500€, but both seemed to suffer from creative fatigue at roughly the same rate. How is that possible?!🫠
 
bungabunga said:
@DeC
I feel you. Have exact the same experience. Meta is so smoth and TT screams right after I clicked on "publish" for new creatives.😭
My graphics department can't keep up with it all.

Sometimes I feel that TikTok doesn't always tell us the truth. For example, I noticed that I had two campaigns with the same number of creatives but different daily budgets. One was set at 250€ and the other at 500€, but both seemed to suffer from creative fatigue at roughly the same rate. How is that possible?!🫠
Click to expand...
because creatives can take more or less to get fatigue depending on CTR and other engagement metrics. The budget does NOT determine how fast your ads get fatigue.
 
davamosk said:
because creatives can take more or less to get fatigue depending on CTR and other engagement metrics. The budget does NOT determine how fast your ads get fatigue.
Click to expand...
Is fatigue not about shown impressions to users? If not I dont get the system when the same creative has high ctr and then somehow it drops but I have only reached a small percentage of my audience so far.
 
That actually makes sense... in my specific case, my ads were UGC format that had done very very well organically (millions of views), meaning that CTR is high, cheap CPM. Add a high budget on top of that and you have a draining machine. The ads were printing for a week, but now they're claiming back what they initially gave me 😩...

It feels like the safer approach is to not throw too much daily budget in order to be able to keep up with TikTok's pace.
 
bungabunga said:
Is fatigue not about shown impressions to users? If not I dont get the system when the same creative has high ctr and then somehow it drops but I have only reached a small percentage of my audience so far.
Click to expand...
Don't get me wrong - obviously by default, if you have $500/day, you will get fatigue faster than if you have $100 but my point was that a creative can last more or less without fatigue depending on other metrics, not the number of impressions served.

There are some creatives which get fatigue after 3 days and other that can last 7 with a good performance and that is determined by engagement metrics (ctr,hook rate, hold rate, avg watch time, post engagement,etc)
 
davamosk said:
I put 2 ad groups only i have enough budget to generate around 10 events per day - otherwise i do the rotation within the same BAU ad group. As you said, even winning creatives get fatigue pretty quickly on TT so it's a matter of time that you simply have to exchange that winner for a completely new video.

I rotate every time that my asset get more than $100 of spend (although this minimum depends on how much is your CPA in average). What I do is to take the winning asset performance as a reference and if after spending enough money to generate 2-5 conversions, performance is not close to the reference, I just pause the asset and keep rotating.

The higher budget, the higher velocity rotating because more ads will get the minimum
Click to expand...
For those using Smart+ campaigns, do you apply the same testing/BAU rotation logic with separate ad groups, or do you rely on Smart+'s own optimization and just rotate assets within a single campaign? Curious if anyone has found a structured framework that works specifically for Smart+ rather than manual campaigns
 
evrenbozmasi said:
For those using Smart+ campaigns, do you apply the same testing/BAU rotation logic with separate ad groups, or do you rely on Smart+'s own optimization and just rotate assets within a single campaign? Curious if anyone has found a structured framework that works specifically for Smart+ rather than manual campaigns
Click to expand...
I tried both ways and any outperformed the other
 
You must log in or register to reply here.
Back
Top