Budget Allocation Across Meta, Google, and TikTok
A framework for distributing spend based on incremental ROAS, creative fatigue, and audience overlap. Not the “even split” advice you'll find elsewhere.
Chapter 1: The Static Budget Trap
Most brands set their channel budget at the beginning of the quarter and don't touch it until the next planning cycle. “60% Meta, 30% Google, 10% TikTok” gets written on a slide deck and becomes gospel for three months. This is exactly wrong.
The optimal allocation between channels changes weekly - sometimes daily. Creative fatigue on Meta might shift your best marginal dollar to Google. A competitor pulling TikTok spend might open up cheap inventory. Seasonal demand patterns hit different channels at different times. A static budget ignores all of this, leaving an average of 18% efficiency on the table.
18%
Static Budget Waste
Average efficiency loss
Weekly
Optimal Rebalancing
Not quarterly
<5%
Brands That Rebalance
Do it properly
Chapter 2: Understanding Diminishing Returns
Every advertising channel follows an S-curve of diminishing returns. The first $10K on Meta might generate 4x ROAS. The next $10K generates 3x. The next $10K generates 2.2x. At some point, the marginal dollar on Meta generates less return than the marginal dollar on Google - and that's where you should shift spend.
The problem is that the point of diminishing returns is different for every channel, changes over time, and isn't visible in platform dashboards. Platforms show you average ROAS, not marginal ROAS. Average ROAS can look great even when your marginal dollar is generating negative returns - because past efficient spend masks current inefficiency.
Budget Allocation Simulator
Adjust allocation to see how diminishing returns affect blended iROAS.
Blended Incremental ROAS
2.6x
Based on $100K total monthly spend
Simplified model showing diminishing returns. Actual curves vary by vertical, audience, and creative quality.
Marginal vs Average ROAS
Chapter 3: The Allocation Framework
Here's the framework Sam uses to determine optimal allocation. It runs continuously, but you can apply the same logic manually on a weekly basis:
Start with corrected data
Use Parker's correction factors, not platform-reported ROAS. If you're allocating based on inflated numbers, you're optimizing for the wrong thing.
Calculate marginal ROAS per channel
Look at the incremental ROAS of your most recent spend increase (or decrease) on each channel. This tells you the actual return of your marginal dollar.
Equalize marginal returns
The optimal allocation is where the marginal iROAS is equal across all channels. If Meta's marginal iROAS is 2.8x and Google's is 1.9x, shift money from Google to Meta until they equalize.
Apply constraints
Account for minimum viable spend (you need enough on each channel to clear the learning phase), creative availability, and strategic considerations (brand presence, new channel testing).
Monitor and rebalance weekly
Creative fatigue, competitive shifts, and seasonal changes mean the optimal point shifts constantly. Check marginal returns weekly and rebalance when the gap between channels exceeds 15%.
Chapter 4: Cross-Channel Effects
Channels don't operate in isolation. When you increase Meta prospecting spend, branded search volume on Google goes up. When you pause TikTok, you lose the awareness effect that was feeding your retargeting funnel. These cross-channel effects are invisible in single-channel analysis but critically important for allocation.
Meta → Google Branded
+8-12% branded searchMeta prospecting drives brand awareness. Increasing Meta spend by 20% typically increases branded Google search volume by 8-12%. If you attribute those conversions to Google, you undervalue Meta.
TikTok → Full Funnel
10-15% halo effectTikTok awareness creates demand that converts across all channels. Pausing TikTok often causes a delayed 10-15% drop in total conversions that shows up in Meta and Google, not TikTok.
Audience Overlap
15-25% audience overlapMeta and TikTok audiences overlap significantly in younger demographics. Increasing spend on both creates frequency fatigue faster than either alone. Account for overlap when setting combined budgets.
Chapter 5: Sam in Action
Everything in this guide is what Sam runs continuously. Sam ingests corrected attribution data from Parker, forecasts from Felix, and your business constraints to simulate thousands of allocation scenarios and recommend the optimal budget split.
What Sam Does, Continuously
Calculates marginal iROAS per channel using Parker's corrected attribution data
Runs Monte Carlo simulations (1,000+ scenarios) to find optimal allocation under uncertainty
Accounts for cross-channel interaction effects (Meta → Google halo, audience overlap)
Recommends weekly rebalancing when marginal returns diverge by more than 15%
Simulates the impact of budget changes before you make them - 'what if I shift $10K from Google to Meta?'
Respects your constraints: minimum channel spend, CAC caps, strategic channel requirements
This entire methodology is what Sam runs 24/7 on your data. Dynamic allocation that responds to creative fatigue, competitive shifts, and diminishing returns in real time. No more quarterly budget slides. Just continuous optimization based on where your marginal dollar actually works hardest.
