Your best-performing ad is dying right now. Not dramatically, not all at once, but in the quiet way that makes it invisible until your CFO asks why CAC payback stretched from 8 months to 14.
NP Digital's recent analysis of companies running Google and Meta ads tracked CTR decay week over week, and the curve is steeper than most marketing teams model. By week four, CTR drops 10.8% from baseline. By week eight, you're looking at a 27.1% decline. That's not noise. That's a structural shift in campaign economics that compounds into every downstream metric your board cares about.
The Decay Curve Nobody Budgets For
Most B2B marketing teams plan creative refreshes quarterly. The data says that's roughly three times too slow.
The week-over-week decay pattern from NP Digital's dataset shows the damage accumulating faster than intuition suggests. Week one holds steady. Week two shows a 3.4% dip. Week three accelerates to 3.9%. By week four, you've lost 7.3% cumulative. The curve steepens from there: 10.8% at week five, 15.7% at week six, 24.4% at week seven, and 27.1% by week eight.
What makes this particularly painful for B2B is the interaction with longer sales cycles. If your average deal takes 90 days to close, an ad that fatigues in 30 days is generating impressions for two months after it stopped being effective. You're paying for reach that no longer converts, and the attribution lag means you won't see the damage in pipeline metrics until the quarter is already lost.
Performance marketing analysts note that Meta ads now hit creative fatigue after 3 to 5 days of active delivery, with CTR typically dropping 20 to 40% from peak by day seven. The platforms have gotten more aggressive about recycling impressions to the same users, which accelerates the decay curve beyond what historical benchmarks would predict.
The CAC Payback Math
Here's where the CFO conversation gets uncomfortable. Current benchmarks show median B2B SaaS CAC at $702 for self-serve and $11,400 for sales-led motions. The target CAC payback period has tightened to 12 months in the post-ZIRP environment, down from the 18 to 24 month tolerance that prevailed through 2022.
When CTR decays 27% over eight weeks, your effective CPC rises proportionally. If you were acquiring customers at a 10-month payback, that same campaign now runs at 13 to 14 months. You've crossed the line from "healthy unit economics" to "explain this to the board."
The compounding effect is what kills you. Meta's auction system factors in estimated action rates when determining delivery costs. As engagement signals decline, your ad becomes less competitive, which pushes CPM higher. You end up paying more to reach fewer people who are less likely to convert. One analysis describes this as the "monotony tax": advertisers who fail to provide format variety pay higher CPMs and achieve lower reach than competitors running diverse creative libraries.
What the Refresh Cadence Actually Looks Like
The conventional wisdom says refresh creatives every 3 to 4 weeks. The data suggests that's a ceiling, not a floor.
Channel-specific guidance breaks down like this: TikTok and Snapchat require weekly refreshes due to their content velocity. Meta Feed ads typically hold for 2 to 4 weeks before returns diminish, though some can extend to 6 weeks. Reels fatigue faster, in the 5 to 10 day range. Stories run 10 to 20 days. Pinterest, with its longer consideration cycles, can sustain creative for months.
For B2B specifically, the refresh cadence depends heavily on audience size and daily spend. One practitioner's framework ties creative input to impression run rate: if you're spending £1,000 monthly on Meta with a £15 CPM targeting 37 million users, you'll serve 66,600 impressions monthly. The math says it would take 555 months to reach your entire audience once. But that's not how targeting actually works. Your true addressable audience is smaller, which means frequency builds faster than the raw numbers suggest.
The operational implication: remarketing campaigns need more aggressive refresh cycles than prospecting. Frequencies exceeding 10 per week can work for lower-consideration products, but only if you're rotating 5 to 10 creatives concurrently and swapping in 2 to 3 new versions weekly.
Detection Before the Damage
The warning signs show up in the data before they show up in pipeline. The challenge is catching them early enough to act.
The clearest signal is when CTR decreases while CPM remains consistent. That pattern means your targeting is still correct, but your creative has lost its pull. The inverse pattern, rising CPM with falling CTR, indicates audience oversaturation from repeated exposure.
ROAS plateauing or declining without changes to media strategy is another indicator. When return on ad spend drops with consistent budget and targeting, the messaging has stopped resonating. Frequency metrics help, but high frequency alone isn't diagnostic. The problem emerges when high frequency coincides with stagnating or declining performance.
AI-powered detection systems now monitor fatigue indicators hourly, analyzing patterns that manual review would miss or catch too late. The claim is that AI can predict fatigue 24 to 48 hours before it becomes visible in standard reports, compared to the 7 to 14 day lag typical of weekly or monthly manual monitoring.
The Budget Allocation Question
Current benchmarks show paid media absorbing roughly 30.6% of the average B2B marketing budget, making it the single largest discretionary line item for most teams. Digital channels overall now take 61.1% of total marketing spend.
The question isn't whether to allocate budget to creative refresh. It's how much of your paid media budget should be reserved for testing and iteration versus scaling proven winners.
One framework suggests holding 10 to 20% of budget for testing new channels and experiments. AI-driven ad tools can cut acquisition costs by 20 to 30% when applied across existing campaigns with disciplined testing. The savings come from faster creative iteration cycles, with top performers testing 47 ads per month versus 11 for laggards, and better predictive targeting on platforms like Meta Advantage+ and Google Performance Max.
Brands running AI-generated creative tested through algorithmic experimentation report a median 14% paid CAC reduction year over year, with the top decile reporting 28%. The gap between AI-mature and AI-laggard advertisers is now larger than the gap between any two paid channels.
The Pilot Plan
If your current creative refresh cadence is quarterly, here's a 3-week pilot to pressure-test the decay curve in your own data:
Week 1: Baseline your current campaigns. Pull CTR, CPC, and frequency data by creative asset for the past 8 weeks. Map the decay curve for your top 5 performers. Identify which assets have been running longest without refresh.
Week 2: Implement frequency caps on your highest-spend campaigns. Frequency capping limits how often the same person sees your ad, reducing fatigue and extending creative lifespan. Set caps at 3 to 5 impressions per user per week for prospecting, higher for remarketing.
Week 3: Launch 3 to 5 creative variants against your current top performer. Test different hooks, not just different visuals. Measure CTR delta at day 3, day 7, and day 14. Document which variants hold performance longest.
The goal isn't to prove that ad fatigue exists. It's to quantify the decay rate specific to your audience, your spend level, and your creative approach. That number becomes the input for your refresh cadence SOP and your creative production budget.
The Forecast Implication
Ad fatigue isn't a creative problem. It's a forecasting problem.
If your pipeline model assumes stable CAC and you're running the same creative for 8 weeks, you're building a forecast on a foundation that erodes 27% over that period. The gap between projected and actual pipeline won't show up until deals fail to close at the expected rate, by which point you've already committed the budget.
The fix is building decay assumptions into your paid media forecast. Model CTR decline at 3 to 4% per week after week two. Adjust CPC projections accordingly. Budget for creative production as a percentage of media spend, not as a separate line item that gets cut when efficiency pressure hits.
Your CFO will appreciate the math. Your pipeline will appreciate the realism.
