A self-custody wallet at 3.4× ROAS across MENA and SEA
The app was buying installs and losing users before funding. We rebuilt tracking around the first funded transfer and cut every source that could not produce one.
Client: Crypto wallet app (confidential)
The challenge
Activation was the bottleneck, not installs. Users opened the wallet once and never funded it, and install-to-funded-transfer ratios varied wildly by source. Without sub-ID visibility the team could not tell which inventory was worth buying.
The approach
Push and in-page push were split into separate campaigns per market, with S2S postbacks wired to the first funded transfer. Creative was rebuilt around three concrete angles: fee-free transfers, local currency support and a two-minute setup. Sources that produced installs without funding were cut within 48 hours. Creative was written per market rather than machine-translated.
What we would do differently
No campaign is perfect. In this account we would start source-level cuts one week earlier and ship local-language creative for the second-largest GEO before scaling spend. Both would have shortened the path to the target cost per event.
The difference was tracking we could act on. We finally knew which source funded wallets and which one just delivered installs.
Let’s build your version of this.
We will map the offer to the formats and sources most likely to produce your target event.
Other campaigns.
First-time depositors for an iGaming brand in LATAM
A new operator needed FTDs in three markets where CPMs were climbing. We rebuilt the funnel around deposit quality…
A broker’s first 1,000 funded accounts
A regulated broker wanted funded accounts, not leads. We combined native and interstitial with a qualification-first…