Your Meta Ads ROAS Is Declining. Here's What the Data Actually Says.

If your Meta Ads ROAS has dropped 20-40% over the past 18 months and you are blaming your creative team, stop. The problem is structural, and it is hitting almost every advertiser on the platform.

Meta CPMs have increased roughly 30% since Q1 2024. The average CPM for US-targeted campaigns hit $14.80 in Q4 2025, up from $10.20 two years prior. Meanwhile, click-through rates have stayed flat or declined slightly. More expensive impressions, same engagement. The math gets ugly fast.

But the real story is not just rising costs. It is what Meta's own reporting hides from you.

The CPM Trend Is Not Slowing Down

Three forces are pushing Meta CPMs higher in 2026. First, advertiser competition. Meta's ad inventory has grown modestly (Reels, Threads), but advertiser demand has grown faster. More buyers bidding on the same eyeballs means higher prices.

Second, Advantage+ campaigns. Meta has pushed most advertisers into its automated campaign types, which optimize for conversion volume at whatever cost. That means the algorithm is happy to pay $18 CPMs if it thinks the user will convert. Great for Meta's revenue, rough on your blended efficiency.

Third, audience saturation in mature markets. If you are targeting US/UK/CA, the addressable audience has barely grown in three years. You are fighting for the same pool of users, and your frequency numbers prove it.

Creative Fatigue Is Real, But It Is Not the Whole Story

Yes, creative fatigue matters. Average ad creative lifespan on Meta has dropped to about 10-14 days before performance degrades. In 2021, you could run a winning creative for 30+ days. That means you need 2-3x more creative assets just to maintain the same output.

But here is what most performance marketers miss: even with fresh creative, your frequency is climbing. The median frequency for conversion-optimized campaigns is now 4.2x per week in the US. That was 2.8x in early 2024. You are showing ads to the same people more often, and each additional impression delivers less.

What Meta Reports vs. What Actually Happened

This is where it gets interesting. Meta's Ads Manager routinely over-reports ROAS by 20-50% compared to what marketing mix modeling shows. Why? Because Meta counts view-through conversions aggressively. Someone sees your ad, does not click, and buys 6 days later through Google Search. Meta claims that conversion.

When we run MMM on accounts spending $100K-$500K/month on Meta, the typical finding is that Meta's true incremental ROAS is 30-40% lower than what Ads Manager shows. For some brands, Meta is still profitable at that adjusted number. For others, it is barely breaking even.

The only way to know which camp you fall into is measurement that does not rely on Meta's own tracking. Check our demo report to see how MMM separates actual Meta contribution from platform over-reporting.

What to Do About Declining Meta ROAS

1. Get an independent measurement baseline.Run MMM to understand Meta's true contribution. If Meta reports 4.2x ROAS but MMM shows 2.6x, your budget decisions should use 2.6x.

2. Check your response curve for saturation. Most advertisers spending $200K+/month on Meta are past the saturation point. The response curve flattens, and every additional dollar returns less. Moving $30K-$50K from Meta to an unsaturated channel (YouTube, programmatic, even podcasts) often improves total ROAS.

3. Rotate creative faster. If you are not producing at least 15-20 net-new ad concepts per month, your creative pipeline is too slow for the current Meta environment.

4. Test Advantage+ Shopping separately. For ecommerce, ASC campaigns often outperform standard setups. But measure them independently because ASC includes retargeting and brand queries, which inflates the apparent ROAS.

The bottom line: Meta is still a top-tier ad platform, but the days of trusting its self-reported numbers are over. Independent measurement is not optional anymore. It is how you find out whether the decline is real or just a reporting artifact.