Neon Bank — Scaling a challenger bank app in 30 days

Challenge

Neon Bank launched in Q4 2025 with a strong brand but no scalable acquisition engine. iOS and Android CPI were both trending above $8, well above the $4 payback threshold their LTV model required. Their internal team had built one Meta campaign and one Apple Search Ads broad-match group — neither tied to MMP post-backs, neither informed by LTV data.

The board approved a 90-day window: hit a $4 blended CPI at 3× install volume or pause paid spend.

Approach

• Rebuilt the Meta creative pipeline around 12 hook variants tested weekly
• Switched Apple Search Ads from broad to exact + discovery with negative sculpting
• ASO refresh on App Store listing — icon A/B test, updated screenshots, keyword depth audit
• Wired Adjust SKAN post-backs to the Meta Advantage+ App Campaigns layer
• Rebuilt the LTV model with weekly cohort updates and a 30-day payback target
• Standing weekly creative review — winning hooks scaled, losing ones killed at day 7

What we shipped

• 48 unique creative concepts in the first 60 days, 6 winners promoted to evergreen
• Three ASA campaign types running in parallel: exact-match, discovery, brand-defensive
• A new App Store listing with 6 fresh screenshots, a video preview, and the keyword set rebuilt around three challenger-bank verticals
• A weekly executive dashboard pulling from Adjust + the in-house BI stack — replaced the stale agency-style PDF report

Results

Within 30 days CPI dropped to $4.60 on iOS and $3.90 on Android at 3× install volume — the board target hit ahead of schedule.

By day 60, blended D7 ROAS reached 3.1× and weekly creative iteration was self-sustaining inside the team.

The system stays in place today: Neon Bank ships a new hook every 10 days, the LTV model refreshes weekly, and the agency hands off learnings each Friday call. Their internal team owns the playbook end-to-end.