Validating a mortgage product for the open market (in two weeks)
The challenge wasn't designing a product. It was designing an experiment that would give leadership enough signal to make a real investment decision.
DISCOVERY
Starting with the people who knew the product best
Before touching any interface, I mapped the existing operational flow with the people closest to it: the customer experience team, credit analysts, and partnership managers. Most of Pontte's home loan leads came through partner referrals: real estate consultants and construction company backers. Understanding that funnel revealed what a B2C version needed to replicate, and what it could safely skip.
I interviewed five partners: financial consultants, construction company backoffice staff, and independent real estate agents. The key question wasn't "do you know our product", it was "where does the handoff break down?" The answers pointed to lead submission as the friction point, not product awareness.
Project requirements
-
Minimal use of the tech team
-
Use only campaign testing funds
-
​Release the page in two weeks
-
Increase the number of leads for the product
STRATEGY & EXECUTION
The experiment design
The existing acquisition flow was relationship-driven (partners submitted leads manually). For a B2C test, we needed a channel that required minimal tech infrastructure, matched consumer behavior, and was trackable enough to make the data actionable. WhatsApp was already in the mix for partner communication. We used it as the conversion endpoint for the campaign page, with proper UTM tagging to attribute traffic by channel.
To build and ship the website, I chose Webflow over coded alternatives specifically because it supported embedded scripts, essential for campaign tracking. No-code wasn't a compromise; it was the right tool for the experiment's constraints.
Trade-off navigated
A fully branded, polished product page would have taken 6–8 weeks. We optimized for speed of learning, not production quality, and framed the page explicitly as a campaign test, not a product launch.
What we deliberately left out
No calculator, no full eligibility flow, no account creation. The goal was qualified interest, not end-to-end conversion. Adding complexity would have muddied the signal we were trying to read.
IMPACT AND WHAT IT UNLOCKED
The experiment answered the question
R$ 6 Million
in home mortgage leads generated through the campaign
+81% visitors
increase in unique visitors site-wide (halo effect on Home Equity)
B2C market fit
Results unlocked engineering budget and official product page
The results gave leadership the confidence to greenlight an official product page, built with the tech team. The experiment also surfaced CRM gaps (we couldn't handle the lead volume with the existing tooling), becoming the next infrastructure investment. A two-week design test reshaped the product roadmap.
After MVP, with the product page designed and implemented