Q1 delivered a 3.2x blended ROAS on $124,500 total spend — up from 2.8x in Q4. Google Search scaled efficiently. TikTok launched mid-February and beat every prospecting channel on CPA.

The problem is where the budget went. Meta Prospecting consumed 24% of Q1 spend at the worst CPA in the portfolio — $48.20. That is 53% above the blended average. Shift that budget to Google Search or TikTok at Q1 efficiency rates and blended ROAS moves to 3.9x.

Google Search is the anchor. TikTok is the emerging story. Meta Prospecting is the drag. Q2 budget decisions follow that order.

Google Search (4.8x ROAS, $42,000 spend): Brand Exact Match returned 7.0x on $18,200 — branded traffic always converts. Non-Brand High Intent hit 4.0x on $14,800 in competitive keyword territory. The remaining $9,000 in broad match returned 2.4x. Broad match is where budget quietly disappears.

Meta (two different stories): Retargeting delivered 3.1x on $19,400. Prospecting: 1.4x on $29,800. The iOS attribution tightening in late January broke prospecting — cold audience signals degraded. Retargeting held because those users had prior site visits. That gap does not close without a privacy-safe signal strategy.

TikTok (2.9x ROAS, $8,200 spend, 6 weeks): Launched February 14th with three UGC creatives. $31.40 CPA against Meta Prospecting at $48.20. No creative fatigue yet. The ceiling is unknown.

The CPA curve bends downward starting Feb 14. That is not optimization. That is a new channel with fresh creative entering at $31.40 CPA — $17 below Meta Prospecting. The blended number looks like progress. It is not.

Meta Prospecting CPA climbed every week from January through March. $38.40 to $48.20. That trajectory is structural — iOS attribution tightening broke cold-audience signals. Spending more on Meta does not fix a signal problem. It amplifies it.

The question is not whether TikTok works. At $31.40 CPA with no creative fatigue, it works. The question is whether it holds as spend scales from $8,200 to $16,000. Six weeks of data is a promising start, not a guarantee.

Your blended CPA improvement masks the real story. Meta Prospecting is deteriorating. TikTok is compensating. If TikTok efficiency slips at scale, the blended curve reverses — and Meta will not be there to catch it.

24% of budget goes to the worst-performing channel (Meta Prospecting). That is the clearest reallocation signal in the data.

Q2 moves:

• Cut Meta Prospecting from $29,800 to $15,000. Releases $14,800. Meta retargeting stays — it is working.

• Add $8,000 to TikTok. Scale the UGC approach with two new variants. At current efficiency, $16,200 total TikTok spend returns 3x.

• Add $6,800 to Google Search Non-Brand. High-intent campaigns have headroom. Broad match does not — keep flat or reduce.

If these shifts hold Q1 efficiency, blended ROAS moves from 3.2x to 3.7x on the same total budget. An incremental $62,000 in revenue without increasing spend.

Google Search earns 4.8x on $42K. Meta Prospecting returns 1.4x on $29.8K. Same budget tier, 3.4x difference in return. Efficiency is not a function of spend volume — it is a function of channel-audience fit.

The correlation is strong (r=0.91). Spend more, earn more — on average. But averages are where strategy goes to die. Meta Prospecting is the outlier that drags the portfolio below the regression line. Every dollar above the trend line is Google Search. Every dollar below it is Meta Prospecting.

This chart makes the reallocation case without a spreadsheet. Shift the outlier's budget toward the trend. The math is not complicated — the politics might be.

1. Kill or restructure Meta Prospecting Interest Stack. 1.2x ROAS on $15,200 spend. Not a scaling problem — a signal problem. Interest-based targeting on Meta has degraded post-iOS. Replace with a Conversions API feed or pause entirely.

2. Brief two new TikTok creative variants before Q2 starts. The UGC approach works at $31.40 CPA, but a single creative set fatigues by week 8. Plan three active variants rotating. Scale budget to $16,000 for Q2.

3. Audit Google Broad Match for negative keywords. $9,000 in broad match returned 1.7x — 65% below Non-Brand High Intent. The spend is not the issue, the keyword targeting is. A structured negative keyword list recovers 20-30% of wasted spend within 30 days.

No new budget required. Better allocation of what is already being spent.

If the reallocation happens: Meta Prospecting drops to $15,000. TikTok scales to $16,200. Google Search Non-Brand adds $6,800. At Q1 efficiency rates, blended ROAS moves to 3.7x. An extra $62,000 in revenue on the same spend. The CPA curve keeps trending down.

If nothing changes: Meta Prospecting keeps $29,800 at 1.4x. iOS signal degradation continues — Meta prospecting CPA has been rising 2-3% per month. At that rate, Meta Prospecting CPA hits $51 by end of Q2. Blended ROAS drifts from 3.2x toward 2.9x. The Q1 headline becomes a peak, not a floor.

TikTok creative fatigue: The current UGC set has run 6 weeks. TikTok UGC creative fatigues at week 8-10. Without new variants, TikTok CPA rises 15-25% by mid-Q2. With two new variants rotating in, CPA holds near $31. The gap: $16,000 in TikTok spend returning 2.9x vs 2.2x. That is $11,200 in revenue lost to creative fatigue.