What to do when everything's on fire?

[[DeC]]

Not sure if it's just me, but sometimes I just feel like throwing all out through the window when performance dips overnight like crazy on Meta

I'm managing an ad account that spends around $5K a day across several territories. We've had good performance for the last few months. It's true that US is starting to see some ad fatigue, but that was normally compensated with good performance in other countries, specially T1 Europe because we opened these territories months later, meaning that it lags behind in terms of ad creatives.

So far so good until the last couple of days, where performance dipped EVERYWHERE. I know they've had "high disruptions" in ad delivery, ads creation / editing and whatnot. My experience says that when a big outage like this happen, it takes at least a couple of days to recover, but we were already recovering from the last week unacknowledged outage when another one happened.

We have other ad networks open, like Google Ads or TikTok, but most of our ad spend has gone to Meta (like 80%) because it was driving good results and was allowing us to test ads fast enough.

But now I see that even the testing campaign is f**ed. To give some perspective, I'm used to see <$1 CPI on US, because the ad is viral enough, but since during the night, it's speedrun'd the campaign's budget, delivering some _great_ results of >$20 PER INSTALL, which is kind of crazy.

Is it just me or anyone else feeling the same sometimes? How do you cope with this kind of frustration? Is there any safe approach that has worked for you to keep you away from such rollercoasters?

 

[davtech]

Not sure if it's just me, but sometimes I just feel like throwing all out through the window when performance dips overnight like crazy on Meta

I'm managing an ad account that spends around $5K a day across several territories. We've had good performance for the last few months. It's true that US is starting to see some ad fatigue, but that was normally compensated with good performance in other countries, specially T1 Europe because we opened these territories months later, meaning that it lags behind in terms of ad creatives.

So far so good until the last couple of days, where performance dipped EVERYWHERE. I know they've had "high disruptions" in ad delivery, ads creation / editing and whatnot. My experience says that when a big outage like this happen, it takes at least a couple of days to recover, but we were already recovering from the last week unacknowledged outage when another one happened.

We have other ad networks open, like Google Ads or TikTok, but most of our ad spend has gone to Meta (like 80%) because it was driving good results and was allowing us to test ads fast enough.

But now I see that even the testing campaign is f**ed. To give some perspective, I'm used to see <$1 CPI on US, because the ad is viral enough, but since during the night, it's speedrun'd the campaign's budget, delivering some _great_ results of >$20 PER INSTALL, which is kind of crazy.

Is it just me or anyone else feeling the same sometimes? How do you cope with this kind of frustration? Is there any safe approach that has worked for you to keep you away from such rollercoasters?

I totally get the frustration — when you're managing budgets at that scale, these kinds of swings can drive anyone crazy. Unfortunately, with platforms like Meta you have to assume that situations like this will happen from time to time.

One thing that’s important to keep in mind in UA is that there are millions of things that simply don’t depend on you: attribution pipelines, tracking layers, SDKs, platform outages, broken rules in ad networks, delayed postbacks, etc. Sometimes performance drops are caused by something completely outside your control.

What does depend on you is how you approach the problem. Take a breath, step back, and try to isolate the issue into the smallest pieces possible, so you can evaluate them one by one and identify where the problem might actually be coming from.

A few things that have helped me deal with these scenarios:

a) Be mentally and operationally prepared for these situations to happen regularly.
Meta outages, delivery instability, and sudden performance drops are not rare events. Having internal expectations that performance can fluctuate (sometimes for reasons outside your control) makes it easier to react calmly instead of scrambling.

b) Avoid depending too heavily on a single ad network.
You mentioned that most of your spend is on Meta because performance has been strong — which makes sense — but concentration risk is real. Today it’s Meta, tomorrow it could be Google or TikTok. Diversifying spend across networks helps smooth out these kinds of platform-specific disruptions.

c) With the CPI numbers you're mentioning, I would also double-check the fundamentals before assuming it's purely media performance.
If CPIs suddenly jump from <$1 to >$20 overnight, I would start by checking:

1. Have CPMs increased proportionally?
   If CPMs have exploded at the same level (which I doubt), then we could simply be looking at an extreme traffic price increase.
2. What about CPCs?
   If CPCs increased significantly while CPMs stayed relatively stable, that might indicate a CTR problem at the creative level. But based on the numbers you're mentioning, that kind of jump would still feel exaggerated.
3. Are installs actually coming through?
   Do a DoD analysis on install volume and check exactly when installs started dropping. If possible, look at it hour by hour; if not, daily still helps. Then try to correlate any drop in installs with things like:
   
   • recent app releases
   • SDK changes
   • Meta SDK issues (if you're measuring with it)
   • MMP SDK issues (if you're using something like AppsFlyer, Adjust, etc.)

If CPM and CPC don’t justify the CPI spike, it usually means spend is still running but installs are not being attributed properly. That points much more toward a tracking or attribution issue in your data source rather than pure media performance.

It could even be something like Meta not receiving installs it can claim (for example if you're using an MMP and something in the attribution flow broke). Since Meta is an SRN, that can quickly make CPIs look insane.

Share a bit more context if you can and we can try to dissect it together. But above all: take a breath and try not to rely too heavily on a single ad network if that's currently the case.

 

[davamosk]

i've been quite off due to some travelling activities but I am online again and expecting to answer everyone gradually.

In this case, @davtech answer was just perfect, but let me add some points from my side:

1- I'm also seeing overall worse performance in Meta lately because of seasonality in some product. Check your category and assume that seasonality is a factor that simply will affect you from time to time. When that happens, reduce budgets, save money and focus on optimizing other areas like ASO to boost your app traffic as much as you can while costs are unbareable for you

2- Test other platforms - if you are struggling with app campaigns in multiple channels, maybe you can stop and invest that time to build a strong web onboarding that allows you to diversify your marketing activities when costs are just too high

3- check your creative diversification. This is one of the most common issues I see. People leveraging in 1-2 assets for months and they get surprised when the creatives don't work anymore. That's not meta fault, that's simply ad fatigue. Check your CTRs and your engagement rates (reactions/impressions) WoW. If you see a declining trend in both and a inclining trend in CPIs/CPAs, you know where the issue is so stop, and focus on producing new batches of creatives.

And most importantly, don't worry too much.
Life goes way beyond UA and your campaign performance. If you keep working persistently on improving all aspects, you will end up finding a way. Don't let that a few numbers put your mood on the floor and keep pivoting into different strategies with the same energy and motivation