Delivering Results Through Happy Immersion

A Collaborative Environment Focused on Meaning, Growth, and Joy

Alarmy, which has always woken users up completely, has been running non-stop this year to wake them up not just completely, but also feeling refreshed and energized. Early this year, we launched sleep sounds on both OSes to help solve users' trouble falling asleep, creating an exclusive sleep environment unique to Alarmy. Next, we rolled out a sleep analysis feature that quantitatively measured sleep duration to provide feedback. Soon after, we started offering qualitative feedback by analyzing sleep stages (light sleep, REM, deep sleep, etc.) in greater detail. And this quarter, we decided to analyze and provide insights into users' behavioral patterns during sleep, such as snoring, coughing, and talking in their sleep.

That's how the so-called "Snore Collab" was born. Multiple teams joined forces to deliver a refreshing wake-up experience for users faster and more safely. The impact of the "Snore Collab" was truly remarkable. Not only did the value users would get rapidly expand, but the value we gained from working together expanded right along with it. We went beyond merely delivering the product by the due date; we actually got to experience delivering the product in a completely refreshing and energizing way. This means that throughout the entire collaboration period, all members experienced a shared, happy state of flow. How on earth did this happen?

There Were More Reasons Why We Couldn't Do It…?

Truth be told, the Snore Collab was a daunting task that would be difficult to even complete by the deadline.

First of all, there were technical difficulties. We had to determine whether snoring occurred by retrieving various audio values using microphone permissions, and we also had to collect and process additional data (such as snore volume, time, and the actual recording file) to translate it into valuable information for the user. In the process, we had to carefully consider file size and product performance issues as well. Every single one of these wasn't a simple chore, but a highly challenging technical task.

Also, QA was tough. The main use case involves someone actually snoring while sleeping at night, but people generally only sleep once a day, and they might not even snore during that time. That leaves us with roughly 0.8 QA opportunities per day. We tried running QA during the daytime by tweaking various condition values, but that was no walk in the park either.

And there were significant UI/UX changes during the process. While UI/UX tweaks are a dime a dozen when developing a feature, the scope of the changes during this collaboration was so huge that the final design ended up being completely different from what we started building. Since measuring and playing back snoring sounds is a pretty uncommon experience for everyone, the initial product spec was somewhat disconnected from actual usability.

The fact that we had to solve all these difficulties through cross-team collaboration was another hurdle. The snore detection model was handled by DelightRoom's research team, and they were developing it from a physically distant location. Internally, the Mone squad (monetization squad) and the Sleep squad worked together. Just the fact that 8 people had to meet up, align with each other, and collaborate with an external research team demanded a lot of extra time.

The Power of Intrinsic Motivation

The fundamental secret to overcoming all these obstacles was none other than the strong intrinsic motivation of our members. They had intrinsic motivations like meaning, growth, and joy—whether it was aiming to improve specific engineering or design skills, build the capacity to collaborate well with teammates, or just gain experience in making things work out no matter what. Because of this, no matter how many constraints were thrown at them, they cheerfully navigated the process and ultimately achieved the goal.

"That sounds like a pipe dream... Who doesn't know that? But practically speaking, leveraging that is impossible!!"

This might sound a bit idealistic. But that doesn't change the truth. Strong intrinsic motivation was undeniably the sharp-edged secret. So, through this post, I want to look back on our collaboration period and outline the conditions that allowed us to leverage this strong intrinsic motivation.

Creating a Collaborative Environment That Boosts Intrinsic Motivation

If we can elevate positive factors like the meaning of work, growth, and joy while eliminating negative factors like stress and pressure through a few environmental tweaks, the intrinsic motivation of all members will become rock-solid, leading straight into a state of happy immersion.

1. A Pre-Kickoff to Inspire Purpose and Meaning

Before starting the collaboration, we hold a kickoff meeting where all members can discuss the purpose and meaning of the project. We share the ultimate problem we want to solve and its expected impact on the product, but the goal isn't just to share information. It’s crucial to resolve various questions and doubts teammates might have about the collaboration, and to gather a wide range of ideas regarding solutions and collaboration processes in advance. If necessary, these meetings can be held multiple times beforehand. Through this process, everyone feels like they're starting to build the collaboration together, and they solidify their resolve to pull it off.

2. Sprint Planning to Increase Predictability

When executing a large-scale collaboration, the biggest challenge—and simultaneously the most important factor—is increasing the predictability of the sprint. You can maximize predictability for large-scale tasks using the three methods below.

2–1. Sharing a Rough Mid-to-Long-Term Plan

We draw the end picture of the collaboration together on a relatively generous timeline. This is important because looking at the forest—not just the trees right in front of us—provides the motivation needed to run a long race. It also gives us mental breathing room by viewing it as a six-week goal rather than this week's immediate target.

2–2. Setting Goals Broken Down into Sprint Units

Now, we set goals (milestones) broken down into smaller sprint or weekly units. Since we need to set moderately challenging yet realistic goals, it's effective for engineers to divide the work scope through separate, in-depth discussions (like "requirements analysis" and "development architecture").

As a result, we decided to break down the goals for this collaboration week by week, as shown below.

2–3. Securing Adequate Buffers During Planning

Once the milestones are established, a certain amount of predictable surprises can be expected for each. It's essential to secure adequate buffer time to prepare for these foreseeable surprises. Especially for the sprints where you're just breaking ground, it's best to set a generous buffer. Failing to hit the very first milestone ramps up pressure and stress, tanking intrinsic motivation and negatively impacting the chances of hitting subsequent milestones. If no surprises occur and you end up with extra buffer time, that's even better. You can just get a head start on the work for the next milestone.

These three methods are collaboration tactics everyone knows, but not everyone can execute them. When a large-group collaboration kicks off, people usually just start sprinting blindly out of sheer passion and ambition, packing tasks tightly to max out available resources without any real buffer. No matter how strong your intrinsic motivation is at the start, it will quickly run dry under sustained stress. Ironically, taking a relaxed approach without rushing, backed by strong confidence, is the safest and fastest route.

3. Running a Sprint Full of Psychological Safety

Even if you start with a predictable sprint, it's incredibly hard to stop anxiety from creeping in as you go. "Can I really do this?" "What if my teammates suffer because of me?" The important thing here is not to ignore that anxiety, but to honestly acknowledge it and share it with your teammates. You have to share your struggles first so that others feel comfortable sharing theirs. Once a sense of psychological safety is established—"It's okay to make mistakes, it's okay to be wrong. We can just try again"—you can control that creeping anxiety.

Often, this anxiety also stems from an excessive sense of responsibility, but you have to remember that a collaboration isn't a solo mission; it's work done together with your teammates. If you hit a wall, ask your teammates for help. You could swap assigned domains, or tweak the design or product spec. What remains consistently important here is that you solve the problem together, not that you stubbornly stick to the original plan and individual roles.

(Since this collaboration happened to be a snore detection project, we got to listen to each other's snoring quite a bit, and I think that oddly brought us psychologically closer. Everyone honestly asked for help right on time...)

4. The Final Spurt with Clear Boundaries

Towards the end of the collaboration, the big tasks are mostly wrapped up, leaving small, lingering chores scattered across a wide area. This is also when everyone's energy is depleted and impatience hits its peak because the due date is looming. On top of that, multiple changes to the product spec and design have probably left various threads quite messy. It's crucial to do a clean sweep and narrow down the unit goals to be crushed in the remaining time. Among the remaining chores, we quickly sync and then completely ignore anything that won't make the first release. For the tasks that will make it but need some product spec cleanup, we gather them all at once, organize them, and double-check.

Now there's no need to waste energy on unnecessary agonizing. You can pour everything into the remaining unit goals and run a fierce final spurt. By narrowing down the milestones to a daily basis and knocking them out, the finish line starts to come into view.

In short, the process of maintaining high intrinsic motivation during a collaboration goes like this: sync on the purpose and meaning to spark the flame, maximize predictability to let it blaze, boost psychological safety during the sprint to protect it from being snuffed out by stress or pressure, and finally, draw a firm line around the deployment scope at the end to brightly burn the last flames without unnecessary agonizing. Thanks to this, we were able to deliver highly polished results in less time than expected, despite the challenging conditions.

Wrapping Up

Two sprints flew by in a blur of happy immersion. Spending a solid month together, we truly became one team. Because we ran the sprints together and also synced on other tasks outside of the collaboration, it felt even more like a single team.

During the final sprint retrospective, a strange feeling welled up. The fact that we would be doing separate planning starting next sprint already felt awkward. It felt like the meeting room should still be packed with 8 people, and the empty seats would feel lonely. It was also funny to see that everyone felt a similar awkwardness. Unlike our initial meeting, which started out more unfamiliar than expected, we spent such a dense month together that parting ways now felt strange.

With that final retrospective full of bittersweet feelings, our promised collaboration officially wrapped up.

We went beyond merely delivering the product by the due date; we actually got to experience delivering the product in a completely refreshing and energizing way. Throughout the entire collaboration period, all members experienced a shared, happy state of flow.

Just as we experienced happy immersion, I hope our users can feel the expanded value of moving from a complete wake-up to a refreshing wake-up through this new snore detection feature. And through that, I hope they experience happy mornings with Alarmy.