35 overlapping ad sets → 7. +10.5% spend deployed at +31% ROAS.
A major D2C jewelry retailer replaced 35 overlapping retargeting ad sets with 7 stage-and-tier ad sets on the Edgemesh Funnel. Same brand, same creative: 10.5% more spend deployed and Edgemesh-measured ROAS up 31%, with daily volatility down 21%.
Edgemesh AllocationJewelry
Spend deployed
+10.5%
more Meta spend deployed
Realized ROAS
+31%
Edgemesh-measured on Meta
CAC
−31%
~$175 → ~$120
Daily volatility
−21%
1-day ROAS std dev
Matched 6-month windows
Pre-Allocation (Mar-Aug 2025) vs. With Allocation (Sep 2025-Feb 2026)
Metric
Pre (Mar-Aug 2025)
With Allocation (Sep 2025-Feb 2026)
Change
Total Meta ad spend
$2.43M
$2.69M
+10.5%
Attributed revenue (1-day click)
$6.08M
$6.77M
+11.3%
Attributed revenue (7-day click)
$7.83M
$8.84M
+12.9%
Realized ROAS
1.05
1.38
+31%
Customer Acquisition Cost
~$175
~$120
−31%
What changed structurally
35 overlapping retargeting ad sets → 7 stage-and-tier ad sets, mutually exclusive. The brand stopped bidding against itself.
98.8% of user journeys now stay within a single ad set across all five funnel stages. Cross-ad-set handoffs in the lower funnel: ~0%.
Wasted spend on unmatchable users: 24% of Meta spend → 10%.
What it unlocks
Pre-Allocation, the brand was limited by media efficiency: "how much can the channel absorb before ROAS falls off?" Post-Allocation, it was limited only by how much capital they want to deploy.
Daily 1-day ROAS standard deviation fell from 0.91 to 0.72, creating a forecastable account rather than a daily lottery.
End-of-March promo (7-day window): spend up +30%, 1-day click ROAS +21% over the trailing month.
Source: Edgemesh Allocation Case Study, April 2026 · brand name redacted.
What could smarter allocation unlock for your account?
Start with your own historical data and see where capital can be deployed more efficiently.
