Alarm Clock Concept
The following was a design exercise with the following prompt:
Smart Alarm Clock App
• Design an alarm clock app that will retail for $4.99 in the iOS App Store.
• The target audience is health and sleep conscious busy adults ages 25-35.
• The app should have the following features:
  - Single Alarm Support
  - Snooze
  - Come up with one smart or learning feature.
After a quick search, $4.99 is an expensive alarm clock so this thing must have a great experience. Apps that seem to be doing well, and aren't free, have excellent visuals (including live weather) and/or focus on sleep metrics and sleep patterns. My first step is to benchmark and identify great ideas from all other apps (even outside the alarm clock space) and determine how each function of this alarm app could work.
Data Entry:

Apple's default alarm provides a scroll wheel UI to set the time. The default value is the current time. Rounding to the next hour would likely prevent the need to adjust the minute wheel. This is an optimization I implemented in Weave (second screen). A bigger question is the scroll wheels accuracy. The likelihood of several scrolls to get to the right time is high. Using a direct enter system similar to Square Cash would be faster and less work to create an alarm. Optimizations like replacing the "." with "00" would also help.
Dynamic backgrounds and weather based on the time of day are ideas that apps are successfully charging for.
I wanted to take this concept a step further and used Instagram style filters to change a picture meaningful to the user based on the time of day. I believe waking up to the face of a loved one, a majestic landscape, or your pet would delight users.
Another design detail used in this concept is a blur on the image during data entery. This allows the buttons and labels to be the focus of the app during the appropriate times.
When an alarm is set the background also becomes progressively blurry in the two hours leading up to the alarm. This provides an opportunity to use an image mask to present a choice. Wake up and see the meaningful background or snooze and the image remains blurry.
Sleeping Goals:

Sleep metrics are a popular edition to an alarm clock app that user's are paying for. I question the usefulness of detailed graphs since much of the information is not actionable. A user cannot decide to increase or decrease the number of sleep cycles they have. The only lever they have is to adjust the hours of sleep and the question that needs to be asked is "Do you feel rested?"
Instead of focusing on metrics, focus on the action the user can take. Reminders will help the user get sleep. A user can setup a goal to sleep a certain number of hours a night. The app will send then a reminder when they should think about getting ready for bed. Based on the rested answers from the user, the app can learn to adjust this reminder so the user can meet their goals.
The following diagram displays my thought process the lifecycle of this app
Screens showing an alarm being set
Screens showing an alarm set then fast fowarding to the alarm in the morning
Screens to ask if the user had a restful night of sleep?
A required, but simple form (like uber) asks if the user felt rested. This give the app the data to learn and point out behavior that is and isn't meeting your goal.
Screens to show insights from data
A user can see their progress over time and the app can point out trends that may be hurting the progress towards the goal.
Wireframes showing the structure of the app. The main screen in the middle has two navigation options at the top.
Alarm Clock Concept

Alarm Clock Concept

Alarm Clock

Creative Fields