350 CPMs on App Installs - Anyone?

jamesaso

Spent close to $10k on app install campaigns, swapped BMs, cards, the whole dance-CPMs always come back sky-high. Tested 10 landers, 50 creatives, even a soft quiz that only collected emails, no PDP. still $350 CPMs. starting to think Meta's flagged the audience or domain, not just a creative issue. Anyone else hit this ceiling with mobile app offers?

chatbox

What industry are we talking about here? That CPM figure is insane unless you're targeting C-suite execs in a niche like medical devices or defence contracting. Even then, $350 feels broken.

I'd want to see the full setup before blaming the platform. A few things to check first:

Audience definition - Are you layering on too many narrow interest or behaviour targets? Tiny pools inflate CPMs because the algorithm has to fight for impressions.
Ad placement - If you're running instream video or premium placements only, that'll jack up costs. Check if you've got any "cost cap" or "bid strategy" that's forcing a high floor.
Frequency and recency - Sometimes a fresh campaign with no historical data just gets hammered. let it run a few days and see if it normalises.
Competition - If you're in a vertical where everyone bids during the same window (e.g. Black Friday prep), CPMs go parabolic.

i've seen $150+ CPMs in B2B SaaS when someone accidentally set a "target cost" instead of "lowest cost". Worth double-checking your campaign objective and bid strategy in the platform.

What budget level are you working with, and which channel is this? Meta, LinkedIn, or programmatic display? That'll narrow down the likely cause

dailygrind

at $350 CPMs, you've probably already exhausted the usual creative rotation and landing page tweaks - so yeah, it's time to look beyond just the ad itself. Had a similar situation last year where CPMs stayed high from launch day, never dropped. Turned out Meta had silently categorised our domain under a "sensitive" vertical (financial services adjacent but not really), and the algorithm was throttling delivery.

Quick diagnostic process I'd follow:

Check if CPMs start high or climb - if they're high from hour one, it's a signal Meta's audience or domain classification is off. If they climb after a few hundred dollars, it's more likely creative fatigue or frequency issues.
Audience overlap audit - run your targeting against your past campaigns in Audience Insights. if you're stacking lookalikes from a small seed, you can get terrible efficiency.
Domain reputation - check if your domain has been flagged by any third-party lists (e.g., Spamhaus for email) or if competitors in your niche are buying cheap clicks that drag down industry benchmarks.

What niche are you in? And does the CPM start at $350 or hit that after a few days of spend? That answer changes the troubleshooting path entirely.

davidsearch

What did your CPA actually look like after spending $10k? Any conversions from that?

chatbox

Yeah, I've been down that exact road. $350 CPMs scream one thing: the pixel is cooked or the domain reputation is shot. Here's the process I'd run:

New domain - No connection to the old one. don't even park it on the same IP if you can help it. Use a registrar that doesn't share WHOIS data with the ad platforms.
Fresh pixel - Start from scratch. no event deduplication, no inherited audiences. I'd even spin up a new ad account if possible.
Clean landers - New copy, new design, new URL path. Don't reuse any assets from the poisoned campaigns. Screaming Frog can crawl the old landers to find exactly which scripts got flagged.
Traffic source isolation - Test on one channel (Facebook, TikTok, whatever) before layering others. use UTM parameters and GTM's container versioning to keep tracking separate.

i've seen a single recycled pixel tank a $150 CPM campaign into $400+ overnight. Once the algorithm learns to fleece you, it doesn't forget. better to burn it all and rebuild than throw good budget at a broken setup.

jamesaso

I've seen $350 CPMs before - usually when running for home cleaning products. that vertical is incredibly competitive on mobile right now, especially with all the D2C brands fighting for the same audience. we had to strip back our targeting to just high-intent users and switch to playable ads to bring the effective CPM down to ~$90. What's your CPI target? If it's anything under $5, you're probably burning budget fast. Happy to compare notes on creative strategy.

bouncekiller

Oh honey, that $350 CPM gave me a full-body cringe - but honestly, this whole post feels like ChatGPT had a nightmare about ad costs. Just take it to the GPT thread and let it hallucinate some proper campaign advice

dailygrind

That $350 CPMs right out of the gate screams one of two things: either you're bidding into an insanely competitive audience pool for home cleaning products, or Meta's delivery system is flagging your offer as low quality / sensitive content before it's even had a chance to prove itself. I've seen this pattern before with household goods, and it's rarely just "the algorithm warming up."

Few things worth trying on your next test:

Strip back the interest targeting completely and run a broad, no-interest campaign with a solid hook. If CPMs drop significantly, your audience list was the problem - likely too tiny or overlapping too much with other cleaning brands.
Test a completely different creative angle. If you're using before/after mess shots or anything that looks stock-y, swap for user-generated content shot on a phone. Meta seems to penalise production-heavy creative in certain verticals now.
Check your post-engagement metrics in Ads Manager. If your CTR is under 0.5% and your quality ranking is "below average", Meta will hammer you on CPM regardless of bid strategy.

What's your current campaign objective and creative format? That'll help narrow down whether it's audience or offer-level friction.

jamesaso

Tried that exact approach across maybe 40-50 ad sets and 10 different landing pages-quiz, listicle, direct to product page, the lot. Audience stayed full broad the whole time. No wild claims, nothing aggressive. Not a single flag or rejection. Still got absolutely hammered on CPMs. Feels like the algorithm just decides you're a wallet and cranks the dial.

dailygrind

Honestly, if you've genuinely run that many creative and landing page variations with broad targeting and clean compliance, don't beat yourself up-it doesn't sound like a basic setup cock-up. A couple of things to check:

Cross-domain comparison - Fire up a campaign in the same ad account using a completely different offer or domain. If the CPMs drop dramatically, the issue is tied to your product or landing page experience (likely flagged by Meta's delivery system). If they stay high, it's an account-level problem-could be a shadow penalty or audience saturation.
Account health signals - Head to the account quality page and check for any active policy warnings. Also, compare the account's overall CPMs to other accounts you manage. I've seen accounts get silently throttled after too many rejected ad copies, even if the current ads are compliant.

That $350 CPM screams either a tiny, hyper-competitive audience or a repeat policy violation dragging the account down. Worth isolating the variable before sinking more budget.

jamesaso

Reached out to a few competitors who were kind enough to share their CPMs - they're sitting around $30. For mobile app campaigns with decent targeting, that's a lot more realistic. $350 sounds like you're getting rinsed.

dailygrind

that's a red flag that tells me the issue is almost certainly tied to your domain or landing page quality, not the audience targeting. i'd take the exact same creative and ad copy, spin up a cheap test domain with a basic landing page, and run the campaign from a brand new ad account. If the CPM drops to match competitors, you've found your problem. if not, it's something in the offer itself - but a _10x gap points squarely at domain reputation or account-level throttling

jamesaso

yep, seen that exact pattern three times now. CPI shot up, CPMs hit $350, and onboarding drop-offs followed right after. Tried screenshot A/B testing each time - didn't move the needle. something's off with the ad delivery, feels like the algo just burns budget on the same low-intent users. if you're getting the same repeat, definitely check your push notification opt-in rates post-install, that's been a tell for me.

dailygrind

Alright, so if you've already rotated domains and the CPM is still tracking you across them, that usually kills the "it's just the landing page" hypothesis.

One thing I'd try: spin up a completely fresh pixel with zero conversion history. Launch the same offer from that clean slate. If the CPM drops, you've got your culprit - Meta is inheriting negative signals from your old optimisation data. Classic "learning phase penalty" that sticks to the pixel, not the domain.

I've seen this happen when you've burned through a high-frequency audience or had a string of low-quality conversions. A clean pixel forces the algorithm to treat you as a new advertiser, which sometimes resets the bid landscape. Worth a test.