The Wemo app is a smart home app that lets you control lights, plugs, and other household appliances with your phone or your voice (works with Alexa, Google, Apple Homekit).
Overview: We were uniquely identifying a user’s “smart home” using a mixture of SSID, MAC address, and IP address. As manufacturers started shipping routers with duplicate identifiers, this opened us up to a number of security risks. The amount of work required to fix these issues was taking up a growing number of resources. It was effecting our reviews, ability to work with our partners (Amazon, Google, IFTTT), our ability to enter new markets, and our need as a business to provide new features to our users.
Business Goals
We needed to improve how we uniquely identified a Wemo home in order to:
Decrease the cost of maintaining our current home identification method
Expand our opportunities to work with partners (Mainly Amazon, Google, and IFTTT) and enter new markets
Improve security of our products
Improve quality of our products
Access Wemos from anywhere on any mobile device
Make setup easier. A Wemo home can contain Wemos connected to many networks.
The Solution: Wemo needed a truly unique identifier to associate with a home (logical grouping of wemos). After a few different cross-functional workshops with the team, it became clear that linking a Wemo home to 1 or many user provided email addresses would help us to accomplish the above goals.
The Design Challenge: Migrate Wemo users to an account while keeping the above business goals in mind.
My Role: Lead the user experience design of accounts in the Wemo app. Working with the PM team and other stakeholders, I defined the work the UX/UI team would pick up, determined the strategy for how we would work on it, and was responsible for the time it took us to complete it.
Plan the Work Strategically
Since this project touched so many different parts of the app, it became clear that the UX/UI team would need an efficient plan to rapidly produce a lot of designs. A design system would help us do this, and it would help to reduce the work for any future projects. Thus, the Wemo App Design System began with this project. I made it a requirement that where possible, create or reuse a screen pattern or UI component.
Customer Insights
Reviews - users already expect the ability to “log in” to their Wemo home from anywhere. They leave negative reviews when they are on vacation and launch the Wemo app on a new phone. They have to be connected to their local network to gain back access to their Wemos.
Usability testing - When we observe usesr setting up their first wemo in testing, we find that they already have expectations that they should be able to log into Wemo from anywhere and gain access to their Wemos.
Align the Team on High-Level Requirement
Throughout the course of this project, the team became divided on requirements for a number of core features within the Wemo app. I led workshop sessions to look at the pros and cons of different options we had, and compared those options to our business goals.
Define User Scenarios
To help cross-functional teams know what scenarios we were considering and how we would handle them, I’d create visual aids like this truth table.
Frame the Problem
Write stories for other designers to pick up
The Plan
Provide existing users with the opportunity to create an account for a period of time. After a while, require them to make one.
Require new users to create an account before they can set up their first Wemo device.
Other Features
Measurable metrics of success
Define specific, measurable ways of knowing whether we're achieving our business goals with user accounts
Baseline metrics for success
How to track these metrics
Obstacles that may prevent success
A successful account creation experience is as follows
A user is offered the opportunity to create an account
successfully creates one
the account is linked to a logical group,
the group is reconnected to partners (if applicable),
and the email address associated with the account is verified
Learnings
This was our first time using metrics to define the success of an entire project, so I knew the metrics we defined here would be a flag in the sand.
After we launched accounts we saw the following:
40% of our users created an account within 3 months
For existing users:
21% tapped “Create an account” when given the option to create one
15% tapped “Log in” thinking that they already had an account
64% tapped “I’ll do this later”
Since 15% of existing users tapped “Log in” when it’s very unlikely they had an account to log into, updates were made to the flow so that these existing users would be less likely to think they had an account.
Updated flow based on what we learned