How We Found UGLYCASH's Cheapest Path to Funded Accounts Across Latin America
By Kevin Veitia
- Client
- UGLYCASH, a dollar-account banking app for the US and Latin America
- Markets
- Ecuador, Chile, Mexico, Paraguay, Guatemala, El Salvador, Argentina
- Channels
- Meta Ads, TikTok Ads, Google Ads
- Timeframe
- December 2024 to April 2025, and again in 2026
- Goal
- Cost per first funding
The challenge
UGLYCASH gives people in Latin America a US dollar account and card. When we started talking with the team, the company was about two months into go-to-market and needed one thing: active, funded customers.
That distinction matters more in fintech than anywhere. An install is cheap. A sign-up is cheap. A verified customer who actually puts money into the account is what keeps the company alive. Every decision on this account was judged against cost per first funding.
The brief was wide and the budget wasn't. UGLYCASH wanted answers across seven markets at once, on three platforms, with startup money. Then, a few weeks in, Google restricted the account over crypto-related content.
What we did
Graded markets with small, fast tests
We launched country by country with a test budget of about $2,500 per market, then graded each one green, yellow or red. The spread was huge. Cost per install on Meta ranged from $2.69 in Ecuador to $8.42 in Chile, where it worked out to roughly $380 per first funding. Small tests made it obvious where not to scale, before scaling cost real money.
Climbed the optimization ladder
Ad platforms need volume to optimize. Meta wants about 50 conversion events per ad set within a week to get out of its learning phase, and Google's automated bidding needs a steady stream too. Optimize for an event without that volume and you're burning cash on an algorithm that can't learn.
So we optimized as deep in the funnel as the data allowed. Installs first. Then identity verification (KYC) in markets with 50+ KYC events in the last 30 days, starting with Central America. Then first funding, once that event had the volume to support it.
Swapped video for statics
On Meta, static images beat video across every placement, including Reels and Stories, where you'd expect video to win. We moved production to vertical 9:16 statics, which were cheaper and much faster to iterate. That matters on an account where a creative lasted about three to five days before frequency caught up with it.
Refused to gamble the ad account
Google's policy filters flagged about 80% of the video assets as crypto-related, and then restricted the account. A workaround was on the table: a fresh account on a different domain with every crypto reference stripped out.
We told the team not to depend on it. Policy enforcement is algorithmic, and circumventing a restriction is exactly the kind of thing that gets an advertiser banned. A ban can cost you a channel for two or three years of future growth. That's not a trade we'll make on a client's behalf. Budget shifted to Meta and TikTok instead.
Didn't fall for a pretty click-through rate
TikTok looked great on the surface, with click-through rates around 2 to 2.5%. Downstream, it was a different story: roughly $1,000 a day was producing only a handful of visible fundings. High CTR with low-quality users is one of the most expensive illusions in paid social. We cut TikTok budget in the weakest markets and kept it only where it pulled its weight. Here's what the TikTok data showed next.
Read the data honestly, even when it argued against paid
When Google spend paused, organic Android installs went up. That pointed to paid app-store search on the brand name cannibalizing installs UGLYCASH would have won anyway. Our recommendation was to build the mix around TikTok, Meta and influencer drops working together, and to test whether Google campaign types like Display and YouTube avoided the cannibalization branded search caused. An influencer drop in Paraguay, paired with a Meta restart, produced a visible spike in sign-ups.
Results
- A KYC-optimized Meta campaign delivered first fundings at $86 each, versus $151 for a campaign optimized directly for first funding: 43% cheaper
- Cost per install as low as $2.69 on Meta, in Ecuador
- A graded market map and a channel strategy for the next phase
We'll be straight about the headline result: it came from a small sample, about 33 first fundings. It's an early signal, not a law of physics. But the direction makes sense. Optimizing for the deepest event with enough volume taught Meta what a good customer looked like, faster than waiting for the "right" event to reach critical mass.
UGLYCASH came back to us in 2026, when we built a measurement and experimentation framework for its paid acquisition, starting with Meta.
What you can take from this
- Optimize as deep in the funnel as volume allows. A mid-funnel event with volume can beat the 'right' event without it.
- Grade markets with small tests before you scale any of them.
- Test statics before you fund video. They're cheaper, faster, and often win.
- Never risk the ad account to work around a policy. A ban can cost you a channel for years.
- A high click-through rate means nothing if the users don't fund.
- Paid can cannibalize organic. A planned pause is the cheapest way to find out.
Want a second set of eyes on your account? Book a free 30-minute consultation and we'll walk through what's working, what isn't, and where the next win is. Curious what it costs first? Our pricing is public.
Book a Free Consultation