Growth Experiments: Online Banking Within QuickBooks
About the project
QuickBooks at its core is an accounting product that is meant to help a small business owner better manage their finances. Connecting your bank account to QuickBooks is a surefire way of making your life easier because it saves you time and improves the accuracy of your books. The goal of the project was to experiment with ways to get more users to experience the benefits of connecting their bank accounts.
I was the lead Interaction Designer on the project, working with a team of developers, product manager, as well as a behavioral economist for this experiment. Read on to hear about how this test turned out!
The problem and opportunity
Connecting your bank account to QuickBooks is one of the best ways to use the product. But how come not all users were jumping in and connecting their bank accounts? One theory was that the access point to banks was too far down the first use funnel. Another theory was that people may simply not have been able to understand the point of banking and subsequently didn't even try to connect it.
As always, I started the design process just sketching on paper. The goal in the beginning is to go as broad as possible and as quickly as possible so that all the ideas are out there. Then we go narrow by eliminating concepts that wouldn't work in the current state of the product, or that wouldn't work due to some other design rationale. Below is a breadth of the mid-fidelity wireframes that we continued to iterate upon.
For this project, we worked with a behavioral economist who taught us about new ways of presenting ideas to the user.
One concept we explored was to bring up the bank connection immediately during setup. Although we knew that connecting a bank account was the easiest way to work in QBO, we actually asked users to explicitly make a choice and choose the 'better' way of working. The hypothesis was that users would need to actually pause cognitively and make a choice to work the 'automatic' way or the subtly worse 'manual' way.
Another concept we explored was to actually allow the user to experience the 'pain' of tracking expenses manually by letting them enter an expense. All our gut instincts would have said this was exactly opposite of what we actually want the user to do because we knew manually tracking expenses was both slow and prone to errors. The hypothesis was that if the user could experience the 'pain' of manually tracking expenses, they would be more open to trying out a better way of doing things by connecting their bank account.
Our experiment produced some really interesting data. We were able to dramatically increase the amount of initial bank connections. That's great right? Well, the more important indicator of user satisfaction is the amount of people who convert to paid subscribers after their free trial. It turned out that even though we were able to increase bank connections, there was no increase in users who converted to paid subscribers.
This led our team to believe that simply driving more people through the top of the funnel didn't mean much if we didn't fix the experience that we were pushing these users to. So we paused this experiment and turned our attention to the next big project, the bank connect experience.