Optimizing Conversion
How the Lemon Wallet app successfully increased conversion by changing the onboarding flow.
About Lemon
In May 2011, Wences Casares recruited me to join his new startup. At that time, he didn’t have a product or a name, but he wanted to do something in fintech so I packed my bags and moved from Buenos Aires to Palo Alto.
We started developing a basic product that scanned receipts. We extracted all the valuable information and organized it.
With time we learned that the receipt feature alone was not generating enough interest from our users, so we started thinking about changing the product into something more ambitious.
The Product
We noticed people were taking pictures of documents and cards and sending them to their email to have a copy. Why not make a product out of that?
When we launched Lemon Wallet in March 2012, we promoted the product as a “backup of your wallet”, helping to take good pictures of all cards, extracting as much information as possible, and organizing them in a neat way.
Value was mostly perceived too late, often after an incident like a user forgetting their wallet. “I should have scanned all my cards with Lemon” people would say. The magic moment was lost. Note at this time Passbook had not been launched, and skeuomorphism was still a thing.
The Problem
We were having a lot of churn, so we decided to send a survey to our users asking them “Why did you stop using Lemon?” and most responded “I don’t see the benefit of using it”. Users were not seeing the value immediately.
Product in Use
Showing value right after users opened the app could lead to them wanting to have all their cards backed up in case something happened. Directly after seeing that value was the perfect moment to prompt them to take pictures of all their cards.
I worked with Mat Gallipoli to create a series of images to be included in an onboarding slideshow right after the app was opened. That way users could become aware of the different functions of the product that they might not have thought of otherwise.
User Testing
After we had released the new photo onboarding, we noticed the conversions started to improve, but something was still not fully working. We decided to do a series of user tests with a top camera filming the user as they used the app.
First, we discovered users were not thoroughly reading our copy. Second, a surprising number of users were literally rubbing the card on the phone screen, as if it was a scanning machine. We never anticipated that could happen!
Video Solution
We decided to test a video instead of a static photo slideshow. This solved the problem of users not reading the message, and showcased easier to understand live situations. Also, at the very end of the video we included a scene showing the correct way to hold your phone and take a picture of the cards.
Immediately after that frame, the camera opens so users can start scanning their cards right away, after seeing the value, and after being shown how to do it.
We were a little bit skeptical about the video, it seemed a little bit intrusive to automatically play a video right after you open an app, and there was not a lot of information in 2012 about video usage in apps, but we wanted to try it.
Launch
After all those changes we saw a 30% improvement in completion flow. People would scan not only their loyalty cards, or IDs, but also their payment cards. These changes allowed us to reduce churn and move on to the next step, monetization.
Epilogue
Thanks to these improvements and other optimization efforts we were able to launch a very successful paid subscription that led us towards to profitability.
A few months later, LifeLock, an identity theft protection company who was looking for a free product with a paid subscription option, got interested in us.
In December 2013, LifeLock acquired Lemon Wallet. For the following six months, I worked at LifeLock, rebranding the product and adapting our paid subscription option to LifeLock’s new strategy.
me at