Home Redesign for New Customers (First Loan)

The full project is already in production. KYC completion went from 40.1% to 73%, support tickets about the application status are now 5% (before, at least 50%) and the time each commercial operator spends explaining it dropped to 30 minutes (before, 2.3 hours a day).

The Home of first-loan customers did not accompany them on the way to disbursement. The continuation banners (documents, verification and signing) barely generated any interaction: they lived in the area reserved for informational banners and customers did not perceive them as actions. While waiting for disbursement, the Home did not say which step they were on or what was expected of them: only 40.1% completed KYC, it took 2.51 days on average across several sessions and only 7% did it in one. And between the application and the disbursement customers lived in a "black box" that was solved through contact: 54% of promoters handled 5 or more customers a day just to answer "how is my application going?", 75% spent between 3 and more than 5 minutes per customer explaining it because the app did not communicate it on its own, and 60% of the questions were about documents, contract and disbursement ("is everything fine with my documents?", "when will you send me the contract?", "when will you deposit?"). In total 2.3 hours a day were lost and at least 50% of support tickets were about the status.

Three sources. (1) Usability test (November 2024, 13 customers at 2 kiosks) on the previous banner UI: 76.9% success rate, 23.1% drop-off, 25.9% misclick rate, 5.9 min average duration and 4.4/5 satisfaction; customers understood buttons but not banners as action elements, so the problem was the component, not the content. (2) KYC analytics: 40.1% completion, 2.51 days on average across several sessions and 7% in a single session. (3) Internal survey of 102 people (video agents and kiosk staff) and interviews with 12 customers: documents, signing and disbursement concentrate over 80% of the queries, 2 out of 3 critical stages have no visibility and 53% of the team does not know when the customer's contract will arrive.

The Home was redesigned in three parts on a single component system. (1) Action cards: 5 card states that replace the banners, with higher text contrast, an explicit primary button, contextual icons per step (start docs, continue docs, docs in review, document error and contract signing) and copy adjusted to the brand's conversational tone and voice; they separate critical tasks from commercial offers. (2) Dynamic pre-disbursement card: changes with the loan status and shows the approved amount from the first access, the current step and the next one, an explicit CTA when the customer must act and a reassuring message when they don't; it is modular and reusable for other products. (3) Application status tracker: step tracker (Kiosk → Documents → Contract → Disbursement) fed by a single backend endpoint that maps the technical states to 4 visual states (target latency under 200 ms and a 60-second client cache) and also triggers WhatsApp Business messages and Push notifications.

1. Visual architecture communicates intent before content: in the test of the previous UI, customers understood buttons but not banners as actions. The problem was the component, not the copy; separating critical tasks from offers avoids banner blindness.

2. Money is the best CTA: showing the approved amount from the first access is the most motivating element, and the card state must evolve visually (pending → under review → unlocked) so the customer perceives progress. "Waiting" states are as critical as "action required" states.

3. A modular component, not ad-hoc cases: cards with 5 states and configurable slots reduced implementation time, ensured consistency and are ready for new products (Aviva Contigo, memberships).

4. The customer doesn't just want to know their status, they want to know when: the most repeated question was about estimated times. And a single endpoint for three channels (app, WhatsApp and Push) was the architecture decision that made omnichannel design viable.

5. The primary channel must be where the customer is: the non-digitally native customer doesn't open the app proactively, so WhatsApp became the primary channel; WhatsApp Business templates, approved by Meta, are one more design constraint.

1. Take the card system and the application status to the renewal flow: returning customers go through a similar process and currently face the same "black box" (the states change: no KYC and with prior history).

2. Validate the estimated contract date as a displayable data point in the app: it is what the team lacks the most (53% don't know when it will arrive) and it depends on the legal team confirming it can be published.

3. Propagate the action card pattern to other Home contexts that still use informational banners with an implicit CTA.

4. Explore whether the reassuring message during the wait further reduces contacts to support.