The registration process at smava ended at the bank account for payout page. There, we prompted customers to instantly generate a digital bank statement by logging into their bank account. But why? Because it's a win-win-win situation:

• Customers can get personalized credit offers, at better rates, and with less paperwork.
• Banks receive a recent and valid account statement, protecting them from fraud attempts.

• Smava spends less on manual labor like calling people or handling documents, making the company more efficient.

We see the potential in this feature to not only reduce costs and speed up the process for our customers, but also we could pave the way for more digitalization and a eventually a full self-service product.

Unfortunately,less than 10% of our new users in 2023 were using this feature. They felt we were "stripping them down" for no good reason.

Based on the knowledge the team had gathered from previous tests in other sections of our product, we ran a small A/B test focused only on the page’s visuals and communication.

We broke down the content in summarized chunks, and reused UI patterns from other places of our product, making this page look more cohesive with the rest of the journey.

More people showed interest in using the feature, but in the end the number of people completing the process remained low.

This taught us that we could funnel more people in, but our customers aren't either ready or willing to do this.

We followed up with a set of usability tests and interviews, as well as design reviews to gather and collate all available information and find insights.

From the interviews and user tests we confirmed our suspicions that this feature might be too much commitment for our "window-shopping" users, and too obscure for our more motivated users.

We also realized that our documentation was dangerously out of date. We discovered two to three steps in the log in process that we didn't account for.

From the design reviews and competitor analyses we learned that our communication is convoluted and we have accidentally introduced even more steps making portions of the process redundant an inefficient.

After having gone through all the information we had gathered in the previous step, we agreed to do test the following changes:

• Redesign the page to make the bank login the default flow, pushing hard for the bank login
• Removed unnecessary interactions before displaying the bank login, removing 3 interactions
• Displayed by default simple explanations of the process and hid more detailed information in an accordion, uncluttering the page
• Dynamically swapped on-screen information depending on the user's current step, reducing distracting or conflicting information

We followed up with regular competitors research and feedback gathering from our customer consultants. These were also interlaced with user tests, in order to validate hypotheses and understand better the dimension of the risks we will be taking.

Ultimately, the results from these activities should feed into the planning of our AB tests schedules, making sure we weren't having any large idle gaps in our roadmap.

The successful test marked a paradigm shift in the company. Our team proved that the registration route could be more digital, that we can lose some qleads and still bring in more sales. Digitalization brings speed, and also accuracy.

The challenge for our team had become to increase both the number of people doing their bank login, but also keeping qleads at a healthy level. We can't forget that at this point most of our customers aren't using this feature.

After several experiments, losses, and releases, we have a feature that's quite different from its original version - which in AB testing is not always given.

Our route is now more personalized to the customer, communication straightforward, and performing better than before. What's new:

• Placed it on a page of its own, allowing us to keep payout account page focused exclusively on payout information.
• Added redirection timeouts, so customers who aren't able or willing to log-in can bypass this step and get to their credit offers.
• Customized the functionality and communication depending on the individual customer, successfully departing from the one-size-fits-all mindset.
• Reduced the number of interactions by redirecting the customer to their offers after a successful login.
• Successfully integrated this feature into the returning customers flow, increasing qlead and sales volume.

‍