In-app education for Dropbox Business

Increasing user activity by 22% and reducing CX tickets by creating returnable in-app education for Dropbox Business.

Laptop with Dropbox Post Setup Education on it.


1 Designer, 1 Product Manager, 4 Engineers, 1 Analyst, 1 User Researcher, 1 UX Writer.

My role

Product Designer


Apr 2019 - Aug 2019

Website mockup - animation showing the interaction for the returnable education module expanding

The Problem

Data shows that user activity drops to almost zero after the first day in the product. Most users never discover their a-ha” moment.

Qualitative research shows us that even users who have used Dropbox extensively describe the product incorrectly. They are blocked by misunderstanding how Dropbox’s core features function. This problem is a deficiency of the product not the users.

The Solution

Increase users confidence in accomplishing their goals and intelligently suggest best next steps.

My team set out to help users get to their nuanced value by contextually surfacing education. This helped users build the confidence they need to move work forward.

The Approach

We started by running a Design Sprint, which allowed us to quickly identify the largest blocking problems contributing to user drop-off.

We introduced a temporary team mate named Friend Lee to help directionally test whether trying core collaborative features in a low risk way could increase user confidence.

Website mockup - introducing practice team mate

Users are introduced to a practice team mate

Website mockup - showing education help affordance

User are shown where to find dismissed education

Users were first met with a goal-oriented questionnaire to gage user reaction to a personalized onboarding experience.

Website mockup - showing goal centered onboarding

Goal-centric questionnaire

Website mockup - showing differing learning styles starting points

Tailoring education to different learning styles

We introduced practice educational moments to make sharing and access more concrete, helping build confidence in our users. We showed users follow up education clarifying how their actions would affect their team members.

Website mockup - Showing demo mode

Goal-centric questionnaire

Website mockup - showing user action rich confirmation

Tailoring education to different learning styles

Design Sprint Findings

Users had pretty unanimous negative reaction to the Friend Lee character: Stating that it reminded them of Microsoft’s Clippy and all the negative associations they had with that experience.

Most users said they would skip through all of the blocking moments they didn’t associate with setup, wanting to get into the product to start exploring.

Users were intrigued by the information in the supportive educational moments. They appreciated that it made the results of their actions more clear.

Derisking through learning experiments

Before moving forward we wanted to build confidence on how in-app supportive education might resonate with users. We ran a learning experiment to show a module after onboarding that highlighted the most common topics from customer support tickets for users early in their trail.

Website mockup - showing the how-to moduleWebsite mockup - showing the states for the How-to module

The "How-to" module

Iterating to get clearer signal

Coming out of the Sprint we felt some of our design decisions muddied the signal we hoped to get from users. We brought a new prototype to New York to get feedback from a more diverse set of users (rather than always focusing on Bay Area users).

We hoped to get directional feedback on a less blocking experience. We removed personified temporary team member (no more Clippy). We retooled the copy to lead with user value and better clarify the result of their actions.

Website mockup - showing education module expanded state v1

Education Module — Expanded State

Website mockup - Education module minimized tooltip state v1

Education Module — Collapsed State

Second round findings

Participants resonated with the confirmations following their actions. They felt more clear about the outcome and what to do next. They said they would return to the sidebar module if they were stuck, needed clarification or to learn what else they could do.  

Website mockup - showing users action success message

Confirmation UI

Website mockup - showing returnable education module default state v1

Returnable Education Module - V1

The data was clear: create educational moments that are noticeable but don't overtake the user's experience. Users want to be in control and given the option for in-app supportive education.

We empowered uses with different styles by showing short how-to videos, presenting succinct steps to get started, and links to full help articles.

Making a data-informed decision

Results from the learning experiment were positive: we saw more users adding working files to their team shared workspace. More users linking and using the desktop client. And a small but promising predicted revenue lift.

I worked with the UX Writer on my team to make sure the content throughout the experience mapped to users' mental models. I fine-tuned visual design and interactions to give control to users over when and how to access educational content.

Website mockup - showing the final Returnable Education module default state

Returnable Education Module — List View

Website mockup - showing the final Returnable Education module expanded state

Returnable Education Module — Expanded

Website mockup - showing the final Rich Confirmation for user actions

"Rich Confirmations"


Users engaged with the supportive education: 7.6% of users interacted with one or more educational topics in the returnable education module. 29.7% of users saw a rich confirmation and 2.7% clicked the CTA to see the suggested follow up education.

Though revenue impact was relatively flat we saw an increase in engagement metrics around core features and workflows:


desktop client usage


work files added to the team space

The lift in engagement indicated that users were able to build confidence around their shared workspace and became more active throughout their trial. There was also a noticeable correlation between the topics the users most frequently accessed in the in-app education and a reduction in related CX support tickets.

The future of in-app education

The CX Education team at Dropbox is iterating on the education module pattern. Myself, the UX Writer, and the PM on my team are working with CX to ship the next version of this experience to all Dropbox users by the end of 2020.