Summary
Guided Onboarding for Arcade was an experience aimed at guiding users through key features of the software in order to improve the likelihood of retaining trial subscribers within their first 30 days. I was the lead designer tasked with establishing the foundational aspects of this work that would eventually turn into a multi-phased approach. This project would end up shipping to 100% of users and resulted in a significant increase in the team's key activation metrics.
Interaction design
User research
Prototyping
The Problem
Arcade's existing onboarding flow at the time consisted of a front-loaded experience where, after they signed-in, would be required to answer several questions to help Arcade select an appropriate starting kit followed by watching a video explainer outlining how to use Arcade.
72%
of users who make it past the initial survey questions tried to skip the onboarding video that came after the questionnaire
10%
of users who entered into the existing onboarding flow dropped off after 2 initial survey questions
65%
of users who reached the Onboarding video tried to seek and skip to the very end of the video
The Approach
Exploring a guided approach
Based on earlier research insights (see In-App Learning for Arcade), there was evidence to suggest that people generally prefer to learn new techniques or topics related to music by a mix of seeing, hearing, and doing at their own pace.
Observations within the music and non-music tech space
In order to better understand how other products within the tech landscape handled product onboarding, we conducted a light dive into a few products within and outside the music tech space to see if there were any key takeaways we could learn, what were table stakes, where were the opportunities to improve or innovate, and what we should avoid in our approach.
Re-contextualizing onboarding
For this project, I turned extensively to Krystal Higgins excellent book Better Onboarding for examples of good and bad onboarding, as well as best onboarding practices based on existing learnings. It became a manual of sorts that led to a key guiding principle for this project: Onboarding is not a single moment in time, it’s a journey.
Our hunch
Based on what we knew, we developed the following loose hypothesis:
IF we use interactive teaching techniques for onboarding users inside of Arcade, THEN users will be more likely to complete key interactions more successfully which will result in them converting into paid and active subscribers of Arcade
Putting together a loose learning plan
Based on the data and insights we’d gathered so far we had a fair amount of confidence in how we thought onboarding should work within Arcade. However, there were a few things that we wanted to gain more confidence in.
Onboarding should be spread out across the experience
This was a key assumption that we had going in based on the guiding principle outlined earlier. We wanted to ensure that this was going to be a more effective approach than the current experience
Not all onboarding moments require walkthrough guidance
Building guided walkthrough moments could get complex and possibly expensive, so we wanted to better understand if it made sense for every moment to be a guided walkthrough moment
How do current products handle this problem?
We had some indication from our earlier research (see In-App learning) that competitors include learning resources for their products within the app. Do they also front-load their onboarding or do they spread it out in other ways?
Does proficiency make a difference when it comes to onboarding?
How might we ensure that people of all skill levels would benefit from onboarding or if onboarding wasn’t needed, how would we stay out of their way?
Guided vs passive onboarding
Another set of key terms that were informed by Better Onboarding was the concept of “guided” vs “passive” onboarding moments. Based on our earlier research with users and our light competitive analysis, we determined that guided, interactive walkthrough onboarding should be reserved for moments of greater complexity or when key interactions within Arcade, for example playing Arcade’s different key types or using the FX macros. We reserved all other types of onboarding moments as “passive” meaning that the onboarding could be handled by means of text descriptors or even video. The lines could blur in some key moments, such as browsing pages of content within Arcade, or finding sample packs (“kits”).
Outlining the “happy path” for near and future work
In order to help inform the team’s roadmap of work around this space once it was fully staffed, I documented a user flow that outlined the various key moments in time where guided onboarding would be triggered, along with with recommendations for what type of onboarding, either guided or passive, should be used in those moments.
Outcomes
Based on the performance of the guided onboarding variant vs the control, the team rolled out the new phase of guided onboarding to 100% of users. Here's what the team learned:
2x
Guided onboarding users are 2x more likely to engage with Arcade's key features
2% +
The team saw an increase in 30 day trial conversions for those who went through the guided onboarding experience
8%
The team saw a reduction in Day 1 trial cancelations for guided onboarding users
Looking back...
Overall, this project was successful in that we achieved the goals and results we set out to achieve. However, there are a few things I believe we could have done a little differently.
Spend more time validating key assumptions
While this project had limited time and resources, if I could go back, I would have better established key assumptions that we arrived at on the team and spent more time directly addressing and validating or invalidating them through research
Dive deeper into competitive products and pair with user reactions
If I could go back and do this project again, I would have spent a little more time on the competitive analysis part. Through user interviews we could have compared Arcade’s onboarding experience with a few key competitors to better understand what worked and what didn’t
