Project

Ambie - Designing push notifications

www.ambie.fm

My Role

UX Writer / Content Designer responsible for:

  • Writing push notification copy

  • Mapping triggers and push notification timing strategy

  • Aligning messaging with user value and product behaviour

Problem

User research revealed two key issues:

  • Users knew how frequently music was updated but not when, with 88% “listening out for updates”

  • Engagement with the app was low. 75% of managers relied on “verbal feedback” over the Insights page.

Ambie was working hard in the background but we weren’t telling users when updates and improvements were happening. This led to low engagement in-app, and a perception of the product as static over interactive.

Solution

Reposition Ambie from a "background tool" to an "active partner" by designing notifications that surface hidden value without causing user fatigue. Notifications would:

  • Reveals meaningful insights about music performance, venue by venue

  • Highlight product updates

  • Encourage users to engage in-app by performing certain actions

  • Strengthen trust and sentiment toward the brand

Challenge - identifying key moments to communicate

With team meetings, we decided we wanted notifications to 1) inform and 2) drive specific behaviours. Our first challenge was prioritising which notifications to send and when, based on the actions we wanted users to take and which insights were most valuable to them.

Solution

We carried out surveys with users to find out which insights they were most interested in. Research showed they valued music and playlist updates, site-by-site feedback, and information about repeats.

We split potential notifications into two categories: triggered by us and triggered by them. Then, rather than sending notifications at timed intervals, we identified high-value moments where they would feel useful rather than intrusive:

Notifications triggered by Ambie

  • When Ambie updates a user’s library

  • Monthly performance summaries

  • Actions users could take to optimise their experience

  • Keep notifications to a maximum of 1–2 notifications per week, unless user-triggered

Notifications triggered by the user

  • When a “Team” user “likes” or a song or playlist

  • When a “Team” “Manager” or “Admin” submits a playlist request

  • When a playlist hasn’t been used in (x) time

  • When a user over schedules a playlist for its length

I made a note not to notify users when

  • They chose a playlist to play in a venue (they know what action they’ve taken, it’s the impact that’s important) 

  • They “disliked” a song (we already had a toast notification telling them “we won’t play this again”.

  • They built a new schedule (it’s not a material change to playback unless they set that schedule to play).

Challenge - turning backend activity into user-facing value

We discussed how to communicate our background processes into user-friendly updates that users would understand and care about.

Content strategy

  • Rephrase system activity as user outcomes, not internal processes

  • Focus on what changed for them and why it matters

Example

  • Instead of: “Playlist updated”

  • Suggested copy: “Your new tunes are ready! Here’s (150) fresh tracks now in the mix. The user can then immediately review the playlist and feedback via that CTA.

Writing for engagement 

Push notifications can compete for attention and we wanted to avoid them feeling disruptive or being overused.

Content decisions

  • We kept messages concise and scannable

  • Used action-oriented language

  • Used emojis to break up copy and in keeping with Ambie’s friendly, informal voice

Voice

  • Playful but not irreverent

  • Informal but not unprofessional

  • Friendly but not overfamiliar

Challenge - designing role-based messages

During team meetings, we agreed that we wanted each user type to receive updates relevant to their position in the business. I approached the content by drafting messages specific to user types or personas, relevant to their level of responsibility e.g.

  • Location-level notifications to “Teams” and “Managers”

  • Monthly performance updates and business-level notifications to “Admins”

By showing updates relevant to what each user controls, we would make the product feel intuitive, build user trust and serve their specific needs.

Collaboration & Validation

Although we had team meetings to scope this project, to save time I mocked the content and wireframes independently as part of a personal project. To take this from concept to launch, I would have worked with product, data, and marketing teams to ensure the messaging was well-timed and effective.

I’d partner with product to:

  • Define which user actions or system events should trigger notifications

  • Align messaging with product goals

  • Test and validate copy and UI

This would ensure we’re validating our hypothesis and solving the right problems.

I’d work with analytics to:

  • Identify high-value touchpoints (e.g. meaningful changes in performance)

  • Set success metrics such as open rate, CTR, and downstream engagement

  • Analyse behavioural patterns to iterate timing and frequency e.g which notifications actually drive action

I’d work with marketing to:

  • Align on brand voice and tone across channels

  • Ensure consistency between push notifications and in-app messaging

  • Balance engagement with brand perception (avoiding overly promotional language)

  • This would ensure the experience feels cohesive

Validation

Before launch, I’d:

  • Run internal reviews with teams to sense-check clarity and relevance

  • Prototype notification flows and review them in context (timing, frequency, sequencing)

  • Sense-check messaging against example use cases

After launch, I’d:

  • Track key metrics (open rate, CTR, engagement)

  • Monitor drop-offs or disengagement signals

  • Iterate on wording, timing, and triggers based on user behaviour

Outcome

I presented this persona project internally for review. It was received well, but unfortunately not implemented due to competing priorities. Had we rolled it out, I’d expect the impact to be increased engagement, improved understanding of product value, and higher retention from value reinforcement.

More projects