Increasing push notification reach
Project overview
Naked Wines are a direct to consumer online retailer, they operate across a website and app. They wanted to increase the reach of their apps push notifications; here’s how we did it.
When:
November 2023 - January 2023
Platform:
App
Focus:
Push notification activation
Context
I was the the product designer in the app squad. We were a team of 7 including: 1 Product Manager, 2 app developers, 2 BE developers and QA.
Problem
Only a small percentage of customers turn their push notifications on.
Goal
To increase the reach of push notifications on the app by encouraging users to turn their notifications on.
Happy path
Downloads the app
Opens the app for the first time
Sees onboarding flow
Understands the benefits to them for turning on notifications
Turns on notifications
Methods
Comp analysis
Usability testing
User journey mapping
Visual design
Content mapping
Markets
Naked Wines operates across 3 markets
Built and tested in USA, their biggest market
To be rolled out in UK and AUS if successful
Figma
🛠️
Userzoom
🛠️
Miro
🛠️
Mobbin
🛠️
After Effects
🛠️
Figma 🛠️ Userzoom 🛠️ Miro 🛠️ Mobbin 🛠️ After Effects 🛠️
Discovery & research
Understanding the problem
It’s been done before: I reviewed how other companies activate push notifications, to see what we could learn from them. Mobbin is a great tool for learning about a variety of apps, especially those with pay walls.
We can’t expect something for nothing: I did some desk research on various case studies and learned that we can’t expect users to do anything for free. So what could the bargaining chip be?
Where in the app should we ask? After reviewing the possibilities, we agreed that when a user launches the app for the first time, we have the users attention and we wouldn’t be impacting any other important tasks.
What was the feasibility of the desired location? Working with the rest of the squad, we looked into the technical considerations and also reached out to key stakeholders to find out the appetite from the business to develop this area of the app.
Why should I turn it on?
We know from user heuristics that giving someone a reason for doing something makes them more likely to see the benefit and do it. But what constitutes enough of a benefit?
Solution
We started by examining what we could send push notifications about. This varied depending on the model the user had signed up to. One segment of customers benefited from knowing when their free bottle was ready, while another received delivery updates on their upcoming subscription deliveries. We then checked these against previous app user research to ensure they were addressing actual user problems we were aware of.
Solution discovery
We had huge buy-in from the business to improve the onboarding flow and I was keen to create an experience that gets users excited to use the app.
What does good look like? I did a review of onboarding flows from other apps, identifying what best-in-class looked like and where we wanted to position ourselves.
Review of what we’ve got: I conducted a UX review of the current onboarding flow, comparing against what I learnt in the comp analysis.
What’s possible in the app? Often with apps, there’s less restrictions than building for web, so I reached out to the app team to get their input. What could we do? What tech is available and what would they be excited to build? We discussed how to bring the app to life using moving image, haptics and reaction animations.
Are there any technical limitations? Alongside knowing how far we could push our design, it was important to find out what technical considerations there were, like the operating system regulations.
System restrictions
On iOS, you can only trigger the system modal for changing settings once.
Solution
Give the users a reason to want to opt-in by solving actual user problems. I conducted some desk research and ran a workshop with the team to better understand the problem. Focusing on what user problems push notifications could solve, it was important we found solutions for all types of customers.
Design & development
A balance of design and UX
Telling a story to capture attention: I plotted out a content map for my ideas, making sure keep the user at the heart of the story, focussing in on exactly what they get out of using the app and highlighting customer benefits.
Usability testing told us to tone it down: To make sure we created something that excited rather than bored customers, we ran unmoderated usability testing.
Video brought the design to life: I found some B-Roll footage from a photoshoot that was done and stitched together a 20 second reel that told the story of where the customers’ wine came from (see attached).
From storyboard to build: The onboarding flow was interactive and used a lot of animation. To help with dev handover, I created a high fidelity prototype and left detailed handover notes to help the developers bring my vision to life. This worked both ways, as they had specific specs for the videos to minimise the size of the app download.
How long is too long?
We had opposing hypotheses within the team. One thought the shortest possible journey was best for getting more notifications switched on. The counter-argument was that users needed a bit of priming first to set the scene before asking them to enable notifications.
Solution
We conducted a Goldilocks test by creating three flows of differing lengths and put them into an unmoderated monadic user test. We specifically checked which variant made users more likely to switch their notifications on and how they felt about using the app afterwards. In the end, we chose a combination of variants that was short enough to get to the point but still had some of the flair of the other variants.
Results
We saw an increase in users enabling push notifications
End result
We created an onboarding experience that users were excited about whilst increasing the number of users who were enabling notifications.
Key learning
We proved the power of video during this project; users engaged really well with it, significantly increasing their quality perception of Naked Wines.
If I could turn back time
Towards the end of the build, the haptics and animation were trimmed out for the first iteration. It would have been great to prioritise these and plan time savings elsewhere in the build.