AED 6.4M booked revenue · -33% CPA on Google + Meta

A cross-channel paid media account spanning Google, Meta, TikTok and LinkedIn cut blended CAC 51% and grew profitable spend from $140K to $410K/month by migrating every campaign to profit-bidding, deploying Creative Velocity, and rebuilding account structure around margin-tiered objectives. Spend tripled; CAC halved.

The Brutal Problem

The performance director ran $1.7M in quarterly paid spend across four platforms and three product lines. Reported blended ROAS sat at 3.1x; CFO reconciliation against gross margin pegged true contribution at break-even. Every quarterly business review ended with the same conversation: "we need to scale" + "we can't scale without margin" + nobody had a path to both. The director had been promised a 20% spend increase for Q3 if Q2 hit target. Q2 was tracking 14% under. He had been promoted into the role from a senior buyer position fifteen months earlier; the previous performance director had been let go after eight months. He had a similar window. He told us he had stopped sleeping properly in March, was on his third energy drink by 11am most days, and his physician had flagged elevated cortisol at his last physical. His wife had said, the previous Sunday, that she barely recognized him. The paid program had real volume, real channels, and real spend; what it didn't have was an answer to the question "which dollar of this is making us money" because nobody had ever instrumented profit-bidding across all four platforms. The director knew he was the next one out if Q3 missed.

What Made It Worse

Five compounding failures. (1) All four platforms bidding to revenue, not profit — high-revenue low-margin SKUs scaled, opposite cut. (2) PMax shared learning across all product lines, so margin-tier signals were averaged out. (3) Creative refresh was quarterly across all channels — fatigue was structural. (4) Audience overlap between prospecting and retargeting on Meta and TikTok was contaminating learning. (5) LinkedIn (B2B side of the business) was bidding to lead-fills not closed-won, which meant the bidder was rewarding the wrong leads.

The Diagnosis

Quantified per channel. Google Ads: 3.4x reported ROAS, 1.6x POAS-adjusted, top three spend SKUs all bottom-quartile margin. Meta: 2.9x reported, 0.94x POAS, EMQ 4.8. TikTok: 2.4x reported, 0.78x POAS, EMQ 4.1. LinkedIn: $128 CPL, 11% lead-to-opportunity, no closed-won feedback to bidder. Creative fatigue scores red on top-spending assets across all four channels. Audience overlap on Meta retargeting + prospecting estimated at 31% — campaigns were cannibalizing each other's learning. Diagnosis: the account was profitable on platform metrics and unprofitable in reality across every channel, and the cause was identical on each — revenue-bidding without margin signal, plus structural creative fatigue, plus audience contamination. The fix was the same playbook applied four times.

The Solution Stack

12 weeks, 12 steps:

1. Week 1 — Server-side CAPI deployed for Meta, TikTok, Google enhanced conversions, LinkedIn CAPI.
2. Week 2 — Margin feeds. Per-SKU COGS, shipping, return-rate piped into Meta + Google via offline conversion uploads.
3. Week 3 — POAS bidding. All ecommerce campaigns migrated to profit targets. Revenue-ROAS deprecated internally.
4. Week 4 — LinkedIn offline conversions. Closed-won deals piped back with 90-day lookback.
5. Week 5 — PMax margin tiers. Three tiers per product line, learning pools isolated, low-margin SKUs blocked from scaling.
6. Week 6 — Audience hygiene. Meta and TikTok prospecting / retargeting isolated. Lookalike refresh against high-LTV cohorts.
7. Week 7 — Creative Velocity v1. 22 variants/month/channel via brief + creative pod, 72-hour kill window.
8. Week 8 — Geo-holdout incrementality test on Meta to validate true incremental contribution.
9. Week 9 — TikTok Spark Ads deployed for UGC-style creative; CPMs dropped 27% in two weeks.
10. Week 10 — Cross-channel attribution dashboard — single MER / blended POAS view per product line.
11. Week 11 — Spend reallocation v1. Killed five unprofitable campaigns, doubled four undervalued ones.
12. Week 12 — Scaling sprint. Budget raised 80% over three weeks, monitored against POAS floor, with auto-pause guardrails.

The Inflection Point

Week 6. Three full weeks of POAS data had hit Meta and Google, and the Tuesday morning algorithm shift was visible across both platforms. Spend reweighted toward high-margin SKUs the bidders had previously been starving. The director walked into the Wednesday CFO review with a single number — blended POAS for the prior 7 days at 2.9x against a 1.5x floor — and asked for the Q3 spend increase early. He got it that afternoon. By week 10, blended CAC had dropped 51% and monthly profitable spend capacity had tripled. The Q3 target was hit in week 11, with three weeks to spare.

Final Numbers

Metric	Before	After	Change
Monthly profitable spend	$140K	$410K	+193%
Blended POAS	0.96x	2.9x	+202%
Blended CAC	$47	$23	-51%
Meta EMQ	4.8	8.3	+73%
Creative variants / mo	~8	88	+11x
Audience overlap (Meta)	31%	4%	-87%
Contribution margin	0%	+24%	+24 pts

What We Learned / Replicable Playbook

Cross-channel paid scaling fails the same way single-channel does — the bidder bids on the wrong objective. The replicable sequence works identically on Google, Meta, TikTok, and LinkedIn: (1) server-side CAPI / CAPI-equivalent first, (2) margin or deal-value feed to the bidder, (3) profit-objective migration, (4) audience and learning-pool isolation, (5) Creative Velocity at 15-25 variants/channel/month, (6) attribution dashboard against MER and blended POAS, (7) reallocation only after clean signal has two weeks to settle, (8) scaling only against a POAS floor with auto-pause guardrails. Performance teams that try to scale spend before fixing the bidding objective triple their burn rate without moving margin. Sequence is the moat. Speed compounds on top.

Daily-Life Outcome

The performance director slept seven hours a night starting month three. The third energy drink at 11am stopped. The cortisol came down at the next physical. His wife said, in month four, that he was "back." He hired a senior buyer in month five — the first net new headcount on the paid team in two years. He was promoted to VP of growth in month nine. The Q3 spend increase landed. The Q4 one landed twice as large.

FAQ

Does this work for accounts under $50K/month?

Yes — same sequence, faster timeline (typically 8-10 weeks). Volume only matters for the incrementality testing layer.

Can we apply this without TikTok / LinkedIn?

Yes. The sequence is channel-agnostic. Apply it per channel you actually run.

How long until CAC moves visibly?

14-30 days post-CAPI + POAS deployment. Creative Velocity compounds the lift over the 60-90 day window.

Driving services: SEM & Paid Advertising · Server-Side Tracking · DTC Ecommerce Growth. Further reading: POAS vs ROAS · DTC Creative Velocity.

Get a free 48-hour audit → We'll quantify the spend you're burning on revenue-bidding — and the POAS lift you'd see in 30 days of clean profit signal.