Design Sprint! Engaging the Couchsurfing Team on User Activation
--
The Challenge
Being the first in-house product designers for Couchsurfing came with unique challenges. Mobile UX designer Quynh and I had a window of opportunity to define and elevate the role of design-thinking at Couchsurfing.
“Every day is an opportunity to invite your non-UX colleagues into the world of UX . . . as partners in the ongoing project of making your products as user-friendly as possible”
— Leah Buley, “The User Experience Team of One”
More specifically, we had a quarterly focus: how do we increase Activation? (In this case, Activation is the number of users who take an action within 7 days of creating an account).
Turning to past experience, I decided to organize and facilitate a 5-day Design Sprint. To accommodate our diversely-located team, we met in visa-friendly international hot-spot Barcelona.
Setting the Stage
Before arriving in Barcelona, we sent each team member a copy of Jake Knapp’s “Sprint” to read, and shared Activation-related data and research with one another.
We broke into two Design Sprint teams, each working separately and simultaneously. Each team contained community admins, designers, engineers, and project managers. This ensured that diverse viewpoints would be represented.
Liftoff
On day 1 we utilized our existing data to map out user journeys related to Activation. This allowed us to form a common understanding of the problem while
Day 2 saw us putting pencil to paper, separately sketching potential solutions.
On day 3 we shared our sketches anonymously and performed dot-voting: selecting our favorite parts of each sketched solution.
We decided to test out a tailored onboarding experience that connects new users with groups and events based on their interests.
By day 4 it was time to turn what we’d worked on into a prototype.
Building a Prototype “Juuuuuust Right”
In “Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days,” Google Ventures’ Jake Knapp makes the case for a “Goldilocks” prototype: something that isn’t so complex as to tempt attachment by its developers, nor so simplistic that the user will realize it isn’t “real.”
Your prototype needs to find a balance to be juuuuust right. Putting too much work into a prototype will lead to something beautiful — and hard to stay detached from. On the other hand: if your prototype is obviously fake, you won’t get accurate reactions from user testing.
As Knapp aptly puts it: “wasting time on the wrong thing is a major bummer.”
For this Sprint I worked with the engineers on my team to build a prototype using Framer.js. By allowing user input (selection of interests) that customizes future screens, Framer helped us to strike the right balance between function and sunk effort.
Meeting the Customer
Day 5 of the Design Sprint brings users in to test our prototype. A great way to get qualitative data, user testing doubles as a way to introduce engineers to users they otherwise may not have met or properly considered.
Introducing your non-UX teammates (or even your CEO) to users is a fantastic way to build support for user-centered work. The amount of face-time your team has with users directly translates into a stronger product.
We put an ad on Craigslist at the beginning of the week looking for participants who hadn’t used Couchsurfing before. Then, we scheduled times and brought them in to try our prototype.
One team member led each test while the rest observed in real time via screen-sharing. We gave respondents prompts to complete and encouraged them to think-aloud while doing so. All the while we kept a close eye on their actions and body language.
Follow-Ups
Communicating our findings
To get the whole company on the same page, we shared our sprint findings with the wider company and our CEO. To laser-focus on the customer, I created a video compilation of the interviews
Facilitating more community feedback
Keeping momentum on increasing user/developer interaction, we started holding more events where we invited users to try out prototypes (as well as chat, eat pizza, and play board games). I stepped up to run these events and hosted a community picnic at nearby Dolores Park.
Iterative experimentation
To directly follow up on our Sprint experiment, we designed and ran iterative A/B tests on our homepage and in search results. We set KPIs and guardrail metrics before moving forward with each A/B test.
Ultimately, highlighting shared interests on profiles didn’t significantly impact Activation. Through our experiments we did identify our landing page (Explore) as an area with unreached potential, and began conducting follow-up experiments to increase activation and engagement there.
