CASE STUDY - FINTECH / BNPL

How a leading BNPL app cut cost per registration by up to ~41% - across 4 markets, in three iterations.

A disciplined, data-first sprint: it started on a $10k test budget, found the winning hooks and messages, and scaled only what worked - across the US and Europe.

4 markets - 4 languagesCPL goal, compliance-constrained360 creatives3 iterations

-36-41%Cost per registrationCompounded across 3 stages, overall

2.1-4.6%CTR - more than doubledUp from 1.3-1.8% at baseline

4 / 9Hooks validated360 creatives - 4 markets, 4 languages

$10k -> $50kTest budget scaledAfter stage 1 proved out

The client

A leading buy-now-pay-later app running registration campaigns (CPL) across the US and three European markets, each in its own language.

The challenge

A CPL goal, four markets in four languages, and compliance limits on what the ads could say - all to be cracked without burning budget. The team started with a disciplined $10k test; brute-force testing would have spent it long before finding a winner.

THE APPROACH

The Winner Sprint, in three iterations.

Run partly by hand during the fintech data-gathering phase - now automated in the platform.

Cost per registration (indexed)CTR

about -36-41%cost per registration, overall

Find the signal

We built the first creatives from message angles our Market Radar surfaced from fintech ad data, filtered against the team's compliance rules. We validated 4 of 9 hooks and surfaced several winning-message candidates. Cost per registration fell 8% - and on that proof, the team scaled the budget from $10k to $50k.

CPL -8%4 / 9 hooks validatedbudget $10k -> $50k

Lock hooks, work messages and CTAs

Holding the winning hooks fixed, we iterated messages and CTAs. A bigger result: cost per registration dropped a further 20-23% versus iteration 1, depending on the market.

CPL -20 to -23%hooks locked - messages in play

Final assembly and scale

We produced 360 unique creatives across the four markets and ran final tests. Cost per registration fell another 13-17%, country to country.

CPL -13 to -17%360 creatives - 4 markets

Alongside the CPL gains, CTR more than doubled (up to 4.6%). The team walked away with an element-level scorecard: the hook archetypes and message themes that won, and how they shifted from market to market.

WHY IT WORKED

Not more videos.

We started from market data, isolated the highest-leverage element - the hook - first on a small budget, scaled only once it proved out, then layered messages and CTAs on the winners. Each iteration compounded the last.

Want results like this for your app?

Start a pilot

A paid 2-3 week pilot on your own ad spend - lift proven against your baseline.