$33K
in 90 days, week-two decision, not day-one
Popunder + push on tier-1 IT/ES with LATAM expansion. Decisions on week-two reconciled CRM revenue, not day-one ROAS.
A selection of campaigns and switch-network results documented with GEO, vertical, quarter, format, and the actual payout model. No metric disconnected from context. No anonymous quotes. If I don't have a sourced number, I don't publish the slide.

Every case below ships with vertical, GEO, quarter, and the number that matters. Anchored to parallel buys I still run for friends' offers, plus my book at PropellerAds (2018-2023) where the case is older than the date I left. No padded math, no rounded marketing numbers.
$33K
Popunder + push on tier-1 IT/ES with LATAM expansion. Decisions on week-two reconciled CRM revenue, not day-one ROAS.
+62%
Publisher previously saw bid-ceiling CPM, switched to clearing-price reporting. Lift in week one.
4.2x
20+ targeting parameters combined on the same campaign. Carrier filtering caught what GEO alone couldn't.
$18K
Publisher monetized owned social traffic without a site. Fill at 100%, eCPM matched parallel-buy data.
3.8x
Smart CPM cleared the auction on display while CPA Goal chased the conversion on push.
-72%
Anti-fraud at the auction edge, not post-billing dispute. Filtered before the invoice, not after.
+89%
VAST integration with mid-roll fallback. Fill held above 95% across DE peak hours.
$0.18
Server-side validated to CRM. Less than 40% of affiliates run S2S in 2024 — this campaign was in the 40%.
12x
Time-of-day stacked with GEO + carrier. 20+ params, none gated behind a 'premium' tier.
Start
I've watched too many affiliates burn $5K before they figured out the panel was showing them a ceiling, not a clearing price. Start with a tracked test, validate server-side, decide on week-two data. The format is the format. The panel is the difference.