Ding!

Take a look at your phone’s lock screen.

How many notifications are visible?

How do you feel about them? Some are helpful, but some... probably not so much, right? Now... imagine having to figure that out for roughly a million customers.

That’s the challenge my client faced with their app’s push notifications. They had to find the sweet spot — how to engage their customers without driving them away. For them, getting it wrong could cost hundreds of thousands of dollars.

Why does this matter?

Push notifications offer a direct line to customers, but they’re also a double-edged sword. Send too many or the wrong type, and customers can disable them entirely. For my client, this wasn’t just a risk — it was a huge potential loss. So how would they get push notifications right? By listening to their customers.

The brief

At a glance

Key objective

Propose a set of rules to describe which in-app messages are most likely to be seen as valuable by customers and/or least likely to annoy customers.

The client

For reasons of commercial confidence, I won't name or describe my client in this case study.

Project team

UX researcher

Project duration

Three weeks

My client wanted to increase app engagement by using push notifications, but they also wanted to avoid annoying their customers.

Some parts of the business were eager to use notifications for marketing, while others worried this would turn customers off.

I was brought in to help answer two critical questions: What types of notifications do customers find valuable, and which ones annoy them?

The approach

I used qualitative research to dig into how customers really felt. We scheduled 8 interviews—45 minutes each—with both new and long-term customers. Though this might seem like a small sample, the goal was to uncover different attitudes, not consensus.

The insights gathered were meant to form the foundation for a best-practice policy on push notifications—a policy based on real customer feedback, not assumptions.

The interviews

The interviews were conducted via Askable. I crafted a script that focused on eliciting honest, unfiltered feedback about notifications. I asked participants how they felt about different types of messages—some general, some specific.

Interestingly, while I expected stronger reactions to certain notifications, the feedback was more neutral than anticipated. So I started to probe how customers preferred to receive each message: push notifications, emails, or in-app alerts. When I started differentiating the delivery method—new insights emerged.

The analysis

One theme that stood out during analysis was personalisation. Imagine an app sending you a “Happy birthday” message. How does that make you feel..?

Now imagine it says, “Happy birthday, [Your Name]—here’s $5 for a coffee.”

The second message, with its personal touch and added value, received much more positive feedback.

During analysis, I looked for patterns: What made some messages feel valuable and others annoying? Customers appreciated notifications that were specific, timely, actionable, and relational.

Key insights

From the interviews, we learned that participants didn’t mind notifications that provided clear value. For example, they valued messages offering tangible rewards, like a birthday treat or something that could change their day. Notifications which felt more like marketing were tolerated to a point, but once the threshold was crossed, customers were quick to disable them.

Participants drew a clear line between “transactional” messages—those they found useful—and “marketing” messages, which they considered less valuable. Too many of the latter led to notification fatigue and, ultimately, app disengagement.

The recommendations

Based on the findings, I proposed a set of “safety rails” to help my client avoid push notification fatigue:

1. Be valuable: Customers get grumpy if they feel like a push notification is reaching for their wallet. They wanted push notifications to deliver more value to themselves than to the business.
2. Be important: A push notification is an interruption into a customers’ day. They expected the message would be important enough to warrant interrupting their day.
3. Be useful: Customers wanted messages delivered as a push notification to be specific, timely, actionable, and relational. They wanted enough details in the top-level view, that they could then make an informed decision about acting on the message.

By following these principles, the client could maintain customer engagement while avoiding the risk of annoying their customers.

The outcome

I presented these findings to the client and their stakeholders. The research revealed the range and variety of tolerances which customers had toward notifications. It also validated the approach of focusing on personalised, high-value messages. The next step was to scale these qualitative insights into a broader quantitative study to validate the findings and refine the policy further.

The client had invested significantly in push notification technology. Annoying customers via push notifications could have disastrous consequences for the client’s investment. Thanks to UX research, this business can move ahead with a strategy for great customer service.

UX research is more than the warm and fuzzy feelings of delivering great customer service. It’s about making informed, strategic decisions that protect your bottom line.