Frame the problem - and prioritize it on the product roadmap
Form hypothesis from user interviews and analytics
Come up with a strategy to test the hypothesis
Iterate and get user validation on the designs
Incrementally roll out the changes to iOS and Android apps
1 iOS developer + 1 Android developer
Client: New Vida clients didn't completely understand how Vida can help and why they should schedule a video consultation.
Coach: Vida coaches spent way too much time on chat sessions being a sales rep than a coach, sometimes explaining users that they are not chatbots. During initial video consultations, they spent too much time asking basic health questions/stats than on goal setting and planning.
Business: The on-boarding funnel was very leaky: with 60% of app downloads resulting in sign-ups and only 5% scheduling a call. Even with minimum marketing spend, acquisition cost of a new client was very high due to extensive chat sessions and long initial consultation calls.
New Vida users breezed through the on-boarding screens without really understanding Vida's value proposition fully. After signing up, they were dropped into a chat interface with instant messages asking them for available times for a video chat. Many found the experience to be too sudden and some didn't thought this was a chatbot. On the other hand, Vida coaches were frustrated to waste their valuable time on clients who were simply not "ready" for coaching.
We had the following hypothesis to redesign Vida's on-boarding flow:
Communicating Vida's value proposition better will help make coaches time more efficient
Asking questions upfront will not only help better match coaches to the client's goals and availability, but also help coaches get clients who are more "ready" for coaching
EXPERIENCE BEFORE REDESIGN
Understand and Define: I triangulated data from multiple sources (qualitative and quantitative) to understand the problem better. I interviewed 12 coaches to understand what's working and what's not working before, during and after the initial coaching consultation. I also interviewed a few friends and family who tried Vida recently. I also pulled in the funnel metrics from MixPanel.
Diverge & Decide: I moderated a "design studio" session with colleagues from business, coaching, design and engineering. I gave each team a few anonymized chat transcriptions to build empathy with users and coaches. Each cross-functional team was given a research theme topic with the goal of generating as many ideas as possible related to their theme. The group sketched ideas individually and reviewed them together. The design team synthesized the ideas and worked on wireframes.
Prototype & Validate: We built three prototypes in Invision and iterated over them with feedback from users I recruited from Craigslist. We wanted to validate if we should start the on-boarding by asking about goals, which type of coach they wanted or which program they wanted to join. We also wanted to understand how to best ask the questions that will help us match the best coaches.
While the project was very successful in hitting its success metrics, this was a multi-phase project lasting a few months. With the limited resources we had, we could have done a better job defining a much tighter MVP that would have helped us get feedback faster.
We rebranded and redesigned the flow at the same time. While this brought some efficiencies and avoided some duplicative work, this did introduce some feature creep that pushed the deadlines. More importantly, it also made it harder to establish causality of success metrics to branding vs flow changes.
We focused too keenly on our B2C clients. As we started gaining B2B clients, we realized their on-boarding process is very different: it starts with a live coach on the employer premises, and hence the in app on-boarding can be less heavy handed in communicating the value proposition.
Without a good supply of coaches for every region, its hard to match coaches based purely on goals and availability. Going into the interface without an online coach can feel like a dead-end.
Better understand how on-boarding can be tailored based on the type of incoming client - B2C vs B2B.
Some clients still skip the questionnaire: how might we re-engage when clients don't finish questionnaire before the initial consultation.
Continue to iterate on coach matching algorithms.