Why are our Meta CPMs so high, and which changes actually bring them down?
Our answer
First check whether your CPM rose with the market or alone. If alone, refresh creative with new ideas, widen targeting and placements, and fix tracking. Judge every change on cost per purchase, not CPM.
- Confidence
- Medium
- Strongest proof
- 7 Meta rules
- 5 datasets, 7 practitioners, 10 opinions or claims
- Sources
- 19 sources
- 29 statements
- MCPBHDN
New here? Take the 1-minute tour
Why this is the answer
4 reasons and our judgmentMuch of a CPM rise is the market. Meta's own price per ad was 23% higher than a year earlier in the US and Canada in Q3 2026. Triple Whale found the median Meta CPM up 13.24% across more than 40,000 brands (Benchmarking your CPMs and CPAs).
Meta's auction charges less for ads people respond to. The winner has the highest total value, from bid, estimated action rate and ad quality, and a more relevant ad can beat a higher bid. Clickbait and engagement bait do not help (Meta).
Worn-out, look-alike creative and narrow targeting push costs up. Meta marks an ad Creative fatigue at twice the cost per result of your past ads and advises materially different creative. For cost worries it advises a wider audience, not fewer placements.
Gaps in CPM shrink once you count purchases. MHI's video ads cost $27.41 per 1,000 impressions against $19.42 for images, but $63.15 per purchase against $58.66. AdManage found a 1.76 times US to UK CPM gap became 1.11 times in cost per purchase.
Our judgment
Check in order of how cheap each cause is to fix: the market first (Benchmarking your CPMs and CPAs), then account health such as a new account, an unverified business or failed payments (Ramping spend on a new account), then creative fatigue and frequency (What ad frequency is too high), then targeting, placements and ad set count. Skip changes that buy cheaper attention instead of buyers, such as optimizing for reach, clicks or engagement or moving spend to a cheaper format, because cheap impressions did not mean cheap sales in MHI Growth Engine’s and AdManage’s data or in Phil Graham’s and Mariano Petino-Paul’s accounts, and keep images in the mix (The mix of static, video and carousel). Partnership ads may lower CPM, but the support is operators’ reports, so test them on cost per purchase (Creator and Partnership ads).
Our judgment, built from Meta, Segwise, Cody Plofker, Ben Heath, MHI Growth Engine, AdManage.ai, Phil Graham, Mariano Petino-Paul and DTC Newsletter / DTC Podcast: our reading of these sources, not a finding from any one of them.
What the sources say
29 statements from 19 sourcesMeta says an ad set becomes learning limited when it is unlikely to get about 50 optimization events in the week after its last significant edit. Fixes include combining ad sets and raising budget.
“move from purchases to add to cart.”
Charley Tichenor's rule of thumb: CPM and cost per sale up with flat order value is seasonal. Flat CPM with rising cost per sale is your account, and falling CPM with rising cost points to conversion rate and offer.
“If CPM is flat and CPA is up, that's you”
Mariano Petino-Paul says Meta has several times put spend behind new ads that were very engaging but did not sell by the brand's own metrics.
“very engaging, but that won't sell based on the metrics we recognize”
MHI Media's client data shows Instagram CPM at a median $31.15 against $20.30 on Facebook. Video cost more than image per thousand ($27.41 against $19.42) and per purchase ($63.15 against $58.66).
“Advantage+ placements put the money where the cheaper results are.”
Pilothouse says one account's Q4 CPMs fell from a year earlier after it invested in a wider mix of video creative. One account, no figures, causation not tested.
“One Pilothouse account actually saw CPMs decrease year over year during Q4 after investing heavily in a wider mix of video creative.”
Meta says the highest bid or biggest budget does not automatically win its auction: ad quality counts too, informed by over 200 million customer survey responses a year on purchase satisfaction and shipping.
“The highest bid or the biggest budget doesn't automatically win in the auction”
AdManage's Meta data on accounts selling in both markets puts the median US purchase CPA only 1.11x the UK's, against a 1.76x CPM ratio. Observational, on reported purchases, not new-customer CAC.
“reported purchase CPA is not first-time customer CAC.”
Meta no longer lets you pick placements at the ad set level for Engagement, Leads, App promotion and Sales; exclusions move to account level. Meta claims Advantage+ placements cut cost per result 10%.
“you no longer choose placements manually at the ad set level”
Meta's Q3 2026 filing: ad impressions rose 10% and average price per ad 14% worldwide, year over year. In the US and Canada, price per ad rose 23% and impressions 8%.
“We calculate average price per ad as total advertising revenue divided by the number of ads delivered”
Ben Heath says higher CPMs need not mean worse results, since Meta may be better at turning prospects into customers and get the same sales from fewer impressions.
“Meta is now able to do a better job of turning your prospects into customers. So you can get the same amount of sales from fewer ad impressions.”
Meta says an auction is won by the ad with the highest total value, combining bid, estimated action rates and ad quality. It says clickbait and engagement bait do not improve ad performance.
“Engaging in clickbait and engagement bait does not improve ad performance.”
Meta flags 'Creative limited' when an ad's cost per result tops your past ads but is under twice as much, and 'Creative fatigue' at twice or more. It suggests a materially different new ad.
“Keeping your original ad active instead of pausing or turning it off may maximize results”
Meta says ads that are too similar may be treated as the same and shown to the same people, limiting reach and causing Creative limited or fatigue status. It suggests different angles such as price or ease of use.
“When creative is too similar, our system may not recognize it as different and could deliver it to the same sets of audiences.”
Nick Theriot says high CPMs usually mean people don't want to see your ads, from low interest or fatigue, not that Meta is punishing you.
“High CPM is not necessarily Facebook trying to punish you. It's literally because people don't want to see your ads.”
Phil Graham warns that low cost per lead and low CPMs can mislead, citing consulting clients who got many cheap leads that were poor and did not convert.
“I've had people come to me for consulting that got tons of low cost leads that were horrible.”
Cody Plofker says partnership ads with influencers can lower CPMs and lend credibility to new brands.
“partnership ads... they're great for everybody... they can lower CPMs... and then you just get the credibility.”
Relevant Audience, a Facebook ads agency, says Meta costs fall when the ad fits its audience better, not when you cut budget, and comparing cost per result across objectives is meaningless.
“In practice, costs fall when the ad fits its audience better. Cutting the budget does not lower them.”
Ben Heath says failed payments on a Meta ad account lead to higher CPMs and worse results, so add backup payment methods. He gives no data.
“Failed payments lead to higher CPMs and worse results. It's well worth adding back up payment methods.”
Cody Plofker says Winks' new Meta account started near $200 CPMs and fell to $35 to $40 in about two weeks and $18K of spend, soon after verifying its business manager, though he is unsure that caused it.
“a tax you get for being new and not having any signal and losing auctions and just meta not trusting you”
Jacob S. says Meta partnership ads typically get 30% to 50% lower CPMs than other ad formats.
“Typically, the CPMs are going to be a lot better on them.”
PPC Land relays Meta's view that instead of excluding placements you should customize assets per placement, widen the audience for cost and use brand safety controls, which Meta says cannot guarantee compliance.
“can't guarantee that all content and publishers will be compliant”
Triple Whale's benchmark of more than 40,000 brands (Aug 2025 to Jul 2026) puts median Meta CPM at $15.06, up 13.24% from a year earlier, while median cost per acquisition rose 3.14% to $38.99 and ROAS held near 1.88.
Wicked Reports data across hundreds of accounts for July 19-25, 2026: Meta new-customer cost rose 8% to $91 with flat CPC as conversion slipped, a market move rather than creative fatigue.
“The entire increase came from new customer conversion slipping from 2.7% to 2.4%.”
Jon Loomer says restricting audience demographics raises your CPM costs.
“your CPM costs go up.”
Emplifi's Q2 2026 worldwide ecommerce medians show most paid spend going to Facebook feed, 18.42 points more than Instagram feed, with Facebook feed CPC at $0.30 vs $1.30 on Instagram feed.
“Spending on FB feed is 18.42 percentage points higher than on IG feed.”
Segwise, a creative analytics seller, says a $16 CPM is normal for a US ecommerce account, while $40 and climbing signals an account problem. It advises waiting for a 28-day trend before reacting.
“If your CPM is $40 and climbing, something in your account is wrong.”
Balistro, an Indian agency selling Meta ads management, says 2026 CPM rises come mainly from signal loss, and fixing data with server-side tracking beats raising bids. No figures shown.
“The fix is not a bigger bid.”
Blue Sense Digital says hyper-segmenting retargeting raises CPMs and has almost no direct upside.
“That hyper segmentation in retargeting, all it does is increase CPMs.”
Andrew Faris argues a higher CPM on Meta is not an account problem but a sign you are reaching customers who are worth more to you.
“a higher CPM is not actually a problem, what it actually reflects is that you're reaching customers that are more valuable to you.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer