How Carousell launched its first widget for iOS14 (Part 1)
by Julia Maurer & Ha Pham
We set out on a journey to discover, design and build a widget for our iPhone users. The news is not really about the widget itself, as it has existed long before on iOS, but rather about its ability to be added to the home screen on iOS 14. Carousell was launched as a mobile-first classifieds marketplace back in 2012, with the mission to inspire everyone in the world to start buying and selling to make more possible for one another.
Having a widget allows apps like Carousell to provide its users with easier access to core functionalities, and also provides a way to express which apps matter most to them. Here is the story of how we went about it.
(1) Your core user segment is your power users
A widget targets your most frequent users as they will be the only ones that are happy to give up home screen real estate for your app. During user interviews and a segment analysis, we discovered they are mainly sellers that quickly need to respond to updates. The inbox is the most visited section, together with ‘Activity’ tab, our notification centre that keeps users updated about their listings and saved search queries:
Our top power user opens the app 300 times daily. We understand the call for efficiency in getting their daily and most repetitive jobs done, which are:
- “I want to reply to relevant leads as fast as possible”
- “I do not want to forget to answer any leads to maximise my sales.”
- “I want to increase my followers and product likes count to increase trustworthiness and negotiation power.”
Those user needs have been the guiding insight for the development of the widget on iOS 14.
Ask yourself:
What persona/segment best describes your most frequent iOS users?
What are their “Jobs to be done”, where you can reduce friction?
(2) Create a space for creativity: “How might we use the widget”?
We intended to keep the ideation phase as divergent as possible including relevant stakeholders of the organisation. As we all work remotely, we used 15 minutes to brainstorm individually by using an established problem solving framework: “How might we use the widget to support sellers/buyers on Carousell”. Here is a sample of ideas we generated throughout a one hour workshop. No matter which app you work for, you can get inspired by our ideation for your power user base:
How you go about it:
Identify your stakeholders and set up an ideation workshop. Share some initial information, like this video from Apple, before the workshop.
Apply the “How might we” framework
(3) Define what success means for your widget
The success or failure of your widget depends on the metrics that you plan to reach within a defined time period after launch.
The question is, how do we measure an efficiency gain of power users? We used the following metrics to measure success and set a clear threshold that needs to be achieved to define success:
- Increase in inbox/activity tab screen views attributed to the widget from x to y within 1 month after launch, which is a x% increase in screen views.
- Increase the number of chats attributed to widget from x to y within 1 month after launch, which is a x% increase in chats.
- Time to reply to a chat message attributed to the widget from x to y within 1 month after launch, which is a x% reduction in time.
If the feature is a success, we will further invest into this feature or create more widgets. If it turns out to be a failure, we will iterate or even remove the widget.
In the first month, we focus primarily on adoption numbers:
- Adoption of iOS14 among users (this is your maximum addressable market)
- Total number installs
- Penetration of widget installs within the power user segment
- Number of uninstalls
(4) Get inspired by content themes in the market
“Good artists copy, great artists steal”. Before jumping into creating our own widgets, we compiled a mood board of great widgets in the market. We found the following themes:
(5) Set design guidelines for your organisation
Even though we have different product teams responsible for different product goals, our users only experience one Carousell app. Therefore, it is important for all Carousell widgets to tell a consistent story and send a coherent message to our users. As more widgets can be developed by other product teams in the future, we came up with a set of design guidelines as a collaboration tool. The design guidelines allow different product goals to be realised, while maintaining a consistent user-centric point of view, branding and tone of voice. The design guidelines consist of two parts: (1) the design principles for widgets and (2) the UI templates.
The design principles describe the experience we want to create with widgets:
Principle 1: Widgets tell users about new and happy updates about their selling/buying
Our mission is to improve everyone’s life through buying and selling. That extends to the widgets too. Widgets shouldn’t show information users don’t care about (e.g: updates that are too old or not related to selling and buying), or unnecessarily negative updates that make users feel stressful (e.g: “your coins are running out”).
Principle 2: Widgets are sneak peeks. Widget should encourage users to tap to see more information in the app.
The widgets shouldn’t try to be a mini version of the app, or try to replace the need to open the app. It should focus on one value that it can deliver well for our users, such as helping busy merchants to quickly glance if there are important messages they need to tend to right away.
Principle 3: Widgets are personal. They are not meant for broadcasting messages
To deliver the most value to our users in exchange for the real estate on their iOS home screen, we want to be as relevant and personal as possible. The widgets should be like a Carousell personal assistant to our users.
Principle 4: Widgets should recognise achievements & celebrate with users
For our most active users — our merchants, selling could mean a lot of hard work. When we recognise an achievement, such as when a successful sale is made, we should celebrate the success with our users.
The UI templates are flexible patterns that can be used for different product goals
Anatomy
Empty states
It is not exactly exciting to look at an empty widget. We added a little nudge in the copy to encourage our users to use the app more often.
Light & dark mode
Users can choose to have a system-wide Dark Mode on their iOS device, so we make sure that our widgets will fit in. We created a Dark mode version of the widgets. The dark mode widget still follows our brand guidelines, while remaining comfortable to read.
(6) Go through a design iteration process
To release the widget in a sprint, we need to start small and scope down to the first widget you can release. We explored different options for both the small and medium widgets.
Which widget size to choose? We have not yet analysed the impact of different sizes of widgets on our goals, but it should be dependent on your judgement based on how important your app is for the user. Today, a user has around 28 apps on their home screen. A widget will reduce the number of apps on first view by 15–30%, which is a trade-off users need to make. Is your app important enough for them to add your app to the homepage, and if yes, would they trade yours against 4–8 of their other home screen apps?
While the small widget saves screen space for our users, the value it brings to our users is limited. The small widget can only afford one action. There is only enough space to show either the number of updates or one message preview, it feels just like a blown-up version of our in-app red notification dots.
The medium, although taking up more space on the phone screen, gives our target user group — busy merchants — more practical values. It allows merchants to decide if an update is important enough for them to tap on and it gives them the flexibility to go straight to the update, or to go other destinations in the app such as the Inbox or the Activity tab.
(7) Involve product marketing into your Go-to-Market plan
The Carousell widget is a product that is not easy to discover and requires several steps from iOS users before they are able to use the widget. Users need to understand:
- The value of the widget
- How to download iOS14
- How to change their home screen
- How to search and install the widget
This is a rather long journey to onboard a user to a new feature of Carousell. To increase adoption, we recommend a marketing initiative to the relevant audience who are already on iOS14 and hence have lower friction. If a power user opens the app with iOS 14 for the first time, we can encourage this user segment to install the widget.
In the future, there certainly will be iterations required on the marketing plan or the widget itself, we might create more widgets or remove it again. In Part 2, our iOS team will explain how to implement the widget, so stay tuned.
