Digital Wellbeing Initiative • Results

We’ve started to take a closer look at digital well-being and how Niagara Launcher can contribute to improving it. As a first step we wanted to evaluate habits and how people use their phones, so we launched the Digital Wellbeing Initiative, allowing users to willingly share their usage data with us for analysis.

This article presents the insights we’ve gathered thus far.

The Need for Well-Being Tools

To get a first picture of how important digital well-being is, we asked participants of the Digital Wellbeing Initiative how satisfied they were with their phone use. They could select from 1, Very Dissatisfied: "You often feel regret, frustration, or that your usage disrupts your life.", to 6, Very Satisfied: "You feel almost all of your phone use enhances your life, aligns with your intentions, and rarely leads to negative outcomes.".

We grouped these answers into users who are dissatisfied, selecting 1 or 2 (22.10%), moderately satisfied, selecting 3 or 4 (48.32%), and satisfied, selecting 5 or 6 (29.57%). This shows that there is definitive space for improvement in terms of phone satisfaction.

A connection often made with dissatisfaction with phone use is usage duration, which led us to ask the follow-up question, "Do you feel you use your smartphone longer than you intend?".

The vast majority (82.32%) agree with that statement, selecting 4-6, solidifying that most people think they use their phones too much.

Another important aspect to consider is whether users even have a grasp on how long they actually use their devices. So, we showed participants of the initiative their total screen time for the last seven days and asked how they felt about it.

As it turns out, about half (48.6%) did not expect it to be so high. Overall, users felt they were using their phones too much yet were surprised by how much time that actually was.

We think that users need tools to better reflect on their phone usage, which they can use to identify problems and strive for healthier phone habits.

Since users seem unaware of their phone use yet are unsatisfied, we wanted to investigate their phone usage more closely to find ways of helping them.

Phone usage

We think the following dimensions are necessary to paint a detailed picture of overall phone usage: screen time, number of phone sessions and unlocks, session duration, and per-app usage time.

Screen time

One of the most widespread metrics is screen time, or how long your device screen is active. The chart shows the average daily phone duration per user.

It is likely unsurprising that users spend quite a bit of time on their phones, given our interconnected world. However, what might cause us to raise our eyebrows is that the median daily phone use duration is at ~4h 23min, which means half of our users' screen time exceeds this value. Assuming an average of 8 hours of sleep a night (16 hours spent awake), half the users spend more than ~27% of the day with an active phone screen. Applying this ratio to an entire year would mean spending every waking moment from the 1st of January to the 8th of April using your phone.

Prolonged screen time is often considered bad nowadays, but we want to make a distinction, which we will explain later.

Number of sessions

However, looking only at the overall duration of phone use does not paint a complete picture. We don't accumulate all that time in one sitting. We use our phones throughout the day, check the news or new messages, and watch a video here and there. So, how often do we pick up our devices in a day?

Again, half of our users unlock their devices more than 52 times! That's at least once every 18 minutes if you are awake for 16 hours.

The number of sessions could be an indicator of the level of distraction, but the absolute number is missing important timing information. Distraction through phone unlocks will not have a big effect if there are long stretches of time when no unlock happens at all.

Session duration

Now, let's look at how much time we spend on the phone after unlocking it.

In this chart, we can see the cumulative distribution of phone session duration. If we look at duration, e.g., 1 minute, the curve shows us how many percent of sessions are shorter than the duration we are looking at, in this case, 50%. Wait... half of our phone sessions are shorter than 1 minute? We think this is likely due to users checking whether they have new messages, checking the time, or answering that one message real quick.

A quick calculation shows us that we are missing something: approximately 50% of our 52 median unlocks are less than 1 minute long, meaning they contribute a maximum of 52 minutes to our overall phone usage of 4h 23min. We are at least missing another 3h 31min.

This graph confirms our assumption. All sessions shorter than 10 minutes combined account for only around 30% of our overall phone use, while sessions longer than 30 minutes contribute a whopping 40%.

This means tackling those long sessions shows the most promise if we want to reduce screen time.

So, how do we spend time on our phones?

App time

The chart below shows the 15 most-used apps in terms of duration in blue. All other apps are grouped into categories, and the categories are displayed in orange.

What might be surprising is that the combined app usage time of the 15 most used apps is ~51%, while all other apps combined are ~49%. This means most phone usage stems from only a few apps, our most-used ones. Creating specialized features targeting one of these apps would be well worth it. For example, a feature targeting Instagram (5.9%) has more potential than a feature targeting the "Movies & Videos" category (4.7%, excluding YouTube).

Before we updated the Digital Wellbeing Initiative, we had to estimate usage, and important information about lesser-known apps was missing. We used to strip apps with low user numbers entirely, but now we use an elaborate anonymization process to incorporate this data without compromising privacy. This allowed us to reduce the percentage of the Unknown Category from 13.3% to 2.1%.

Conclusion

Now we have a general idea of what phone use looks like. However, we have not yet investigated the relationship between satisfaction and phone usage duration. Is all the time spent on our phones a cause for dissatisfaction?

Is phone use inherently bad?

Strictly speaking, spending more time on your device does not necessarily mean you are not furthering your satisfaction. Using your phone to look at tonight's dinner recipe while cooking, talking with family and friends through video calls, or taking a well-deserved break with your favorite show are all things that likely won't make you feel dissatisfied—the opposite, actually. So, the critical distinction is what you are doing on top of how long you are doing it.

Existing Tooling

Most tools assume that all phone usage is bad or simply group all phone use. Take Android's built-in screen time, for example.

The first thing you are greeted with is today's daily phone usage. Alongside it, you have information about how long you spend in each app, allowing you to narrow down your usage. However, there is no distinction within the usage of a single app, meaning the time you learn How to Make Butter Chicken on YouTube is treated the same as the time you are stuck on YouTube Shorts. Also, there is no distinction between apps that you think have a lot of potentially unwanted usage, like Instagram. This means the task of distinguishing unwanted from wanted phone use falls upon the user.

Unwanted phone use

The overall phone usage duration only gives us an upper bound of unwanted phone use. All phone use could be unwanted, but it's much more likely that only a portion of it is. Our goal is to reduce unwanted phone use.

Which makes us arrive at the core problem: How do we differentiate between wanted and unwanted phone use?

What unwanted phone use means to each person is highly subjective. Some might want to reduce phone use that does not make them happy in the moment (hedonistic approach), while others want to reduce phone use that does not make them feel fulfilled (eudaimonic approach). Since a different definition for each user is impractical, we decided on the following:

This type of use is often characterized by a lack of conscious intention. Users may engage with apps, like social media, out of habit or boredom, only to realize afterward that the activity was unproductive or unfulfilling.

Since identifying unwanted phone use accurately is a difficult task, we approximate unwanted phone usage through potentially unwanted phone use, phone use that is likely to be unwanted or that likely contains a large portion of unwanted phone use.

With this new metric, we can look at phone usage per app again.

Per-app dissatisfaction

We showed users their top 5 most-used apps and their respective usage times. Then, we asked them whether they were happy with how much time they spent on these apps. The chart shows the top 50 most used apps alongside their app category and the percentage of users unhappy with the time spent on them. (Also, the data is from 2021, which is why there are so many third-party Reddit clients.)

The top 8 entries are social media apps, showing that social media apps have a lot of potentially unwanted phone usage duration. Again, we want to stress that not all social media usage is bad; it's just that users are unhappy with how much time they spend there.

Our Next Steps

Better Metric

With all this in mind, we want to focus on measuring potentially unwanted phone use instead of screen time and showing this information to our users. Potentially unwanted phone use is still only an approximation of actual unwanted phone use, meaning there will be room for improvement. Still, we believe this is an essential step in the right direction.

More Insights

We will also expand the Digital Wellbeing Initiative to gain insights into areas where we identified room for improvement.

We plan to do the following:

• Use Android's built-in app usage tracking for better data on per-app usage time
  • ✅ Better tracking is used starting with Niagara Launcher v1.12.11.
• Improve the app categorization for less popular apps so that we can improve our statistics
  • ✅ We managed to reduce unknown apps from 13.3% to 2.1%.
• Add additional metadata to our reports so we can compare results to public research data and investigate cultural influences

For more information, please check out Digital Wellbeing Initiative • What changes?.

New Features

Finally, we plan on testing new features to make you more satisfied with your phone use.

Based on one of our surveys, 91% of users have already tried tools and custom rules to become happier with their phone use. Yet, as shown in the user satisfaction survey, many people are still unsatisfied. This is why we want to experiment with new features to find something that works. Our first attempt at this is the Usage Breaker — a way to be mindful about your phone use.

Participants of the new Digital Wellbeing Initiative will also be able to see whether these features help them personally.

Thanks

Again, massive thanks to those who took part in the Digital Wellbeing Initiative. We could not have done this without you💙

We would appreciate it if you continued participating in the updated Digital Wellbeing Initiative. If you are interested but unsure, please check out our article on the Digital Wellbeing Initiative and our extensive article about what data we collect and how we keep it private.