When user research derails everything: 

As a new start up, onboarding is a crucial part of your product since these screens are the first impression users have of your brand, beyond what marketing efforts achieve. In fact, having a poorly designed user onboarding flow can cost your company a lot, at the very least it costed us a lot. 

Users, having experienced countless onboarding processes with different apps, are often impatient. Relieving this burden creates a stronger, lasting impression of your app. That’s why, while designing the user onboarding flow, we aimed to make our screens as intuitive and seamless as possible, so that users would navigate through them effortlessly. 

To achieve this, we adopted a 'one item per screen' method, where users would focus on one information per screen. For example, instead of entering their name, phone number, and email on one page, users would fill in just their name before moving to the next screen for the next piece of information.

Talking to Customers 

On paper we believed we had a good solution. This type of design is quite preferred by many companies because it is easier than scrolling around and understanding what information to fill in. And us and the stakeholders were satisfied. 

We didn’t even expect that there would be any problem until our first user testing. In our initial user testing we learned that our onboarding process was taking too long. Our testers informed us that they were bored and frustrated because they had no idea when onboarding would end. 

We initially assumed that the reason why our users felt this way was within the nature of the industry we are in—since in finance industry companies have to follow several legislations and report to several institutions. And since these institutions demand us to require some sensitive information from our users such as their identity number, a confirmed cellphone number, name and birthdate; the onboarding tend to be longer.

However, as a neo-bank our process is still a lot shorter than creating a new bank account. 

Moreover, according to our benchmark research most of our competitors had even longer onboarding processes and going through our flow was even faster. As a result we decided on just providing a feedback with a progress bar was simple enough solution.

At the same time, we were also being pressured by the stakeholders, to move onto other features. After all users need to have some other functions after they are onboarded. So, we revised the flow with the feedback bar but did not test it again, as it was a minor update and did not require further validation. We were certain that we fixed the problem and thought if we get churn after this update, there were not more that we could do for users as we were also bounded by law to keep the onboarding that long. 

And boy were we wrong...

In the first weeks after release, we had quite a lot of churn. At this point we still did not know about the problem. Luckily/unluckily, we had a system problem that caused us to shut down our app for 3 hours. And then the comments about how untrustworthy our app was started to pile up. People questioned how could we ask such information, how was that allowed etc. Then it clicked. We were a new "bank", that did not even have a branch anywhere. 

People who just wanted to explore and learn what we were about, were forced to create an account and give a lot of sensitive data. Of course, they didn't. 

And we got 1.9 stars on app store. 

What went wrong

As our initial testers were informed about the brand and what we do, they did not bat an eye on the fact that we need these sensitive data. They were in our office seeing people who created this product. They had no reason to doubt our intentions. For that reason, we took their feedback of "it takes too long to onboard" literally and as the sole reason why real users were churning. 

What we should have done

As we were getting unexpected amount of churns we should have set another research, preferably with the users that signed up. If we had done it, maybe we could have changed our literal interpretation of "it takes too long to onboard" to "people expect to explore the brand first". 

How we fixed it

Since then, we have updated our designs to ease our users onto our brand and provide users the sense of security. 

For that we divided our onboarding to 3 steps. No account, basic onboarding and full onboarding. 
In first step we allow our users to be able to interact with our app without having an account. Basically they can check all the tabs and provided functions. 

On the second step they are creating an account with their phone number and email address only. They are limited with a minimum amount of currency and limited with the functions that they can use. 

And lastly, they can complete their full onboarding and have access to full functions. 

By providing such division, we have created a more trusting interface which users can interact with. 

Closing thoughts

“In all situations where bad design decisions were made, people lacked some information that would have helped them make the right decision.” ~ Jared Spool

The experience of this failure proves how valuable is to talk to your real customers. In hindsight, it was a simple problem to solve. However by not noticing the real issue we only provided a band-aid type of solution to a much bigger scar. And it costed us thousands of users and negative brand reputation. Naturally, it is an important lesson that I will take with me throughout my career.

My key takeaways from this experience are: Even when you think you covered all angles, there might be something you are missing out. So, test your product. Talk to your customers. Always consider the effects of the environment that user test is conducted in. 

To sum up, smallest details can make or break your product. Our onboarding screens have been praised by many industry professionals about its effortlessness before all this happened. However, somewhere in between we forgot that our goal was not to please industry professionals but to inform our users who have no idea about our brand.