From Credit Instrument to Protection Product
Company
Itaú
Position
Lead Product Designer
Product
Limite da conta
Industry
Banking
Platform
Mobile
An automatically activated product that a lot of people wanted to cancel. Limite da Conta is activated when the account is opened, without the customer requesting it. A significant number of customers went straight to the app to cancel it right after opening their account. Product leadership interpreted this as a flow problem and proposed adding friction to the cancellation path. I disagreed with the diagnosis.
The problem wasn't the flow. It was the framing. Customers who canceled shortly after opening their account had a history of debt from revolving credit. This wasn't impulsiveness: it was memory. Making cancellation harder wouldn't change that perception. It would only worsen the experience for people who had a legitimate reason to leave.
"The issue wasn't making cancellation harder. It was changing what the customer understood they were canceling."
The benchmark came from outside the financial sector. I mapped products people keep active without using often, but would never cancel: antivirus software, phone insurance, car insurance. The value of these products isn't in everyday use, it's in being available for the moment everything goes wrong. Limite da Conta was exactly that kind of product, but no one had ever framed it that way.
Customers saw a revolving credit product, a category carrying the historical weight of debt, and reacted with the only instinct that made sense to them: cancel before it was too late.
Before: generic product benefits. After: named active protections that would be deactivated upon cancellation.
Two layers of repositioning within the flow. First, the language: we moved from "Are you sure?" to "Canceling could leave you unprotected," paired with specific, named protections that would be deactivated along with the cancellation.
Second, the structure: canceling the credit limit wasn't just removing a credit product, it was deactivating a set of active protections that would leave the business more exposed.
The full new experience flow: framing as protection, listing active protections, and confirming the permanent cancellation.
"Adding friction to a cancellation flow is a response to a usability problem. The problem here was one of perceived value."
Instead of asking for approval on an idea, I proposed evidence. Internal resistance was understandable: the repositioning was conceptual and hard to evaluate without data. I proposed a one-month A/B test on a representative sample of the customer base. If the metric didn't move, the approach was wrong. Turning a strategic disagreement into a question the data could answer was just as important as any visual decision in the project.
Churn dropped from 0.36% in January to 0.25% in June 2025. When customers truly understand what they're canceling, not a credit product they fear, but a protection they might need, a meaningful share of them decide to stay. Not because the design made leaving harder. Because the design finally explained what was actually at stake.