Over a million developers have joined DZone.

Notifications: An App Love Song to Its Users

DZone's Guide to

Notifications: An App Love Song to Its Users

When it comes to app users, sending notifications at the right time is key, and using AI to create an intelligent layer API will help increase user engagement.

· AI Zone ·
Free Resource

Bias comes in a variety of forms, all of them potentially damaging to the efficacy of your ML algorithm. Read how Alegion's Chief Data Scientist discusses the source of most headlines about AI failures here.

A while ago I read Scott Belsky's post. I loved it. It touched the exact pain point I'm trying to solve every day working for Neura. I even wrote a response.

The smartphone can easily be a black hole of attention -mobile games, Facebook, Instagram, Snap, Twitter - their life line is dependent on multiple app opens a day. They'll nag you with notifications you didn't know existed. From groups you forgot you registered for, to your middle-school friends announcing their latest life events. These apps have zero interest addressing you at the "right time," because all the time is the right time. You're trapped, and unless you reach escape velocity and cancel notifications altogether, you're never getting away.

But, notifications can also help transform lives. Apps from fields like digital health, smart home, wellness, self-improvement, meditation and more can deliver real-life value to their users. When they look at the more successful field, app creators and developers go to what is considered the best practice - notification hell - which is the worst practice for this type of apps.

For example, users of health apps don't want constant reminders that they are not healthy, or that they practice poor self-treatment. They want an app that will remind them in the least intrusive time, that will give them good reasons to open the app and invest time using it.

Belsky covered the solution as he sees it from two angles. In the short term, he says, just handle your notifications better. Invest in better logic. But the example he uses is Slack's logic map for sending a notification which is far from simple. It's a complicated data flow algorithm that was polished in the eight years the company existed (in different names). Still, his claim makes sense. Don't overflow the user, and use analytics to understand the level of engagement you get from each form of messaging.

In the long term, he looks forward to "An Intelligent 'Notification Layer' API" - an API that will optimize notifications using AI. First, no need to look forward, this solution exists today. Second, it's always good news when a product expert acknowledges the product vision of your field.

Our health app user will now have access to an app that responds to its behavior. Just two small examples that will show the difference between "dumb" notifications to one that understands the user's situation:

  • Reminded at wake-up time to take the morning pill instead at a predefined hour.

  • Waiting for the commute to be over and to arrive at their home before notifying to take the afternoon blood glucose measurement.

These are abilities app developers have access to, today. The immediate benefit is an efficient engagement that has a better chance converting to the ROI these apps struggle so much with.

Furthermore, these abilities open a doorway to user segmentation like never before. With the help of AI, the rate of engagement of users with these notifications and with the app features paints a clear picture of how a specific daily routine interacts with an app. These tools can help build contextual content that will resonate with said daily routine.

Engaging users at the right time with exact content that matches their behavior and interest can have the necessary impact that will make sure a product always adapts to its users, creating the allusive usage habit app owners aspire to have with all their users.

Your machine learning project needs enormous amounts of training data to get to a production-ready confidence level. Get a checklist approach to assembling the combination of technology, workforce and project management skills you’ll need to prepare your own training data.

ai ,app design ,notifications ,api

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}