Thierry Design Portfolio - AlfredCam Continuous Recording

Alfred Camera is a mobile application designed to repurposes spare phones as a surveillance camera. The team has faced challenges in supporting various old phone models and has decided to shift their focus towards developing new features and products.

UI Design

UX Design

Illustration Design

2021/08 - 2022/03

CLIENT OVERVIEW

CLIENT

DATE

ROLE

Continuous

recording,

a new feature to bridge service gap

The Story

Originally, Alfred Camera focused on turning users’ idle old phones into home surveillance cameras. However, over the years, the team encountered persistent issues with users using various old phone models, such as connection problems, recording issues, and login difficulties. Therefore, in 2021, the team decided to move forward by developing hardware products to provide a more unified service quality.

Compared to other hardware competitors like Wyze, Xiaomi, and Ring, the current Alfred Camera still lacks an essential feature: long-duration recording. This feature would allow users to continue recording even when offline and could be a crucial solution to Alfred Camera’s longstanding pain points.

the Challenge

Motion Detection Sensitivity

Alfred App has received numerous complaints from users regarding incomplete Motion Detection. In some cases, Motion Detection is triggered, but the recording time is not sufficient, resulting in users only being able to see partial event videos instead of the complete history.
It is important to note that the maximum length of a Motion Detection video is 120 seconds, which is dependent on the movement range and continuity of the event. Moreover, setting the sensitivity too high can result in the capture of unwanted motion videos, such as sunlight changes or falling leaves.

As Motion Detection is commonly used for home security and evidence collection, users demand
high-quality Motion Detection that is precise and complete.

Camera Offline

The issue of camera going offline is one of the most common complaints from users, often occurring due to unstable network connections or outdated devices.

Users in such situations desire a feature that ensures record continuously even when the camera goes offline, with the ability to check for any missed events once the camera reconnects.
This feature would provide users with the assurance that any events occurring during disconnection are recorded and can be viewed when the camera comes back online.

Cloud Service Cost

Our research indicates that 20% of our users account for 90% of our cloud service costs.

These users typically set their camera to record continuously by pointing it at a moving object, such as a fan or a playing monitor.

As a result, Motion Detection is constantly triggered, and these users often generate hundreds of motion videos per day. This hidden risk needs to be addressed to reduce the high cost associated with cloud storage.

Validate the User Demand

There are still a number of questions that require validation:

Can the “Continuous Recording” feature address the needs of existing users? Will it make them more reliant on using Alfred Camera?
Currently, Alfred Camera already has manual recording with Moments and motion detection with Events, each accessible from different locations. Can users easily find the videos recorded with Continuous Recording?
Besides the basic recording and playback functions, what other features do users need?
Generally, what recording duration would satisfy users? If continuous recording is to become a paid feature, what factors would motivate users to pay for it?

IDEAtion

idea 1

Integrate Playback with Live View

The first proposal is to integrate the playback interface for long recordings with Live streaming, allowing users to easily switch between live situations and recorded footage. When users drag the timeline to a specific time, the current live feed seamlessly switches to the corresponding recorded footage. However, this solution requires a stable internet connection for both the camera and the viewing end.

idea 2

Auto-Scaling Timeline Through Gestures

This idea aims to help users find their playback videos more efficiently. When users swipe quickly across the timeline, we interpret it as their intention to locate footage from earlier times. Conversely, if users swipe slowly, we understand that they are searching for a specific moment within that timeframe.

To assist users in finding videos more easily, the timeline will scale dynamically based on their gestures. This feature is especially beneficial for users who record for extended periods, as it allows them to locate their desired footage more efficiently.

Expand

00:00

23:00

22:00

01:00

00:00

23:00

22:00

01:00

Expand

When scrolling through the timeline, and the pointer cross to the previous/next longer scale, and the previous/next time slot will expand automatically.

Experiment process

MVP Version, based on the Proof of Concept version for quick experiment.

Phase 1

Hypothesis: Better retention of new users

by offering Continuous Recording

We selected 1,000 new users with American English devices and divided them into two groups:

one group was given access to Continuous Recording, while the other was not.

We observed the retention rates of these A/B groups to identify any differences.

➤ Result & Learning

After two weeks, 23% of new users chose to participate in the Beta Program, and we observed no significant difference in retention rates between the A/B groups (33.1% vs. 35.3%).

However, it is noteworthy that in our 159 user survey responses, 27.5% mentioned they need at least 24 hours of recording loops.

This led us to consider whether the recording duration might be a key factor affecting user retention. Additionally, some users reported difficulty in pinpointing specific times, making it hard to quickly locate periods with motion detection.

SURVEY

The Feedback Button in the playback interface and the Banner on the camera list are used to gather feedback.

Phase 2

Hypothesis: Improve retention by offering 3-day

recording loop duration.

We selected 4,040 new users from the United States.

2,020 users in Group A were set to have a 3-day recording duration, while the control group (Group B) was limited to a 1-hour recording duration.

We aimed to compare the retention rates between these two groups to see if there was any significant improvement.

➤ Result & Learning

There was no significant difference in retention rates between the A/B groups (77.7% vs. 77.2%).

This suggests that extending the recording duration did not validate our hypothesis of increasing new user retention. However, in a survey of 495 users, 69% expressed that the long-duration recording feature was helpful to them.

Additionally, we found a 55% increase in playback viewing rates among users in the 3-day recording group, likely due to the availability of more footage to watch, resulting in higher product engagement.

Phase 3

Usability Test & User Interview

Due to our inability to understand user reactions solely from data, we conducted a usability test and user interviews.

We selected six highly active Alfred App users who had used Continuous Recording in the past 14 days. We provided them with a test account and asked them to enable the feature and locate playback videos.
Additionally, we requested that they demonstrate their usual way of using long-duration recording.

➤ Result & Learning

A lack of guidance and introduction to new features may be the primary cause of low retention rates.
4 out of the 6 participants stated that long-duration recording is very useful, but they were hindered by insufficient information on its value and usage. Moreover, users often joined the test program without being aware of it, and they tended to skip the “Learn More” button on the program’s introduction page.

Some users mistakenly believed that the Event Book was the only place to view all videos, making it difficult for them to find playback videos.
Finally, users were unsure how to use the timeline features, such as dragging to switch times and scaling the timeline.

optimization

optimization 1

Improving the Timeline for Playback Interface

During our usability testing, we found that participants had difficulty locating specific segments using the timeline. Therefore, we decided to optimize the timeline in the playback interface:

First, we added rewind/fast-forward functionality to the timeline, allowing users to quickly jump back or forward by 10 seconds, a feature not commonly found in our competitors’ products.
Next, we introduced adjustable scaling for the timeline, making it easier to display vertical videos captured in portrait mode. We also supported landscape viewing to allow users to see more details in horizontally captured footage. Finally, we adjusted the timeline scaling to prevent confusion about playback status, ensuring that users can visually perceive changes in the timeline and avoid misunderstandings about video playback.

Each longer scale on the timeline

represents a duration of 60 minutes.

Each longer scale on the timeline

represents a duration of 30 minutes.

OPTIMiZATION 2

Enhancing Guidance for Continuous Recording

We introduced a more detailed tutorial within the playback interface, guiding users through each feature and explaining how they work. When users enable long-duration recording and open the Live interface for the first time, they will see pop-up bubbles that guide them step-by-step through the tutorial. This ensures that users can quickly understand and utilize the various functions of long-duration recording.

OPTIMiZATION 3

Add Date-Selection to Better Integrate with HW Product

As the Alfred Camera product line expands into hardware, such as AlfredCam, users can now record longer videos with long-duration recordings. However, finding specific segments in the playback interface remains challenging for hardware users.

To address this issue, we have introduced a date selection feature with an intuitive calendar design, allowing users to easily locate specific dates, months, and years.

Users can also navigate and browse important moments recorded by AlfredCam more quickly and efficiently.

The Continuous Recording feature is now available to all users with an 8-hour recording duration after optimization.
AlfredCam users can view 8 hours of footage for free, and unlimited footage for premium users, depending on their SD card storage.

takeaway

The Demand for Continous Recording and Its Uncertain Commercial Value
Users need continuous recording, but its potential commercial value remains uncertain.

For users suffering from camera disconnections or incomplete motion detection recordings, continuous recording acts as a safeguard. It effectively addresses current product issues, but our user behavior data reveals that offering continuous recording does not increase user retention rates. Therefore, its potential commercial value is still unclear.

Additionally, our experiments showed that users who use continuous recording are already loyal users of the product. We believe that continuous recording has a selective appeal to this dedicated user base.

Addressing Product Gaps

The lack of long-duration recording could make Alfred Camera less competitive in the hardware camera market. Therefore, continuous recording is a crucial piece in filling the existing product gap.
When users are choosing among various hardware cameras, continuous recording makes AlfredCam more competitive.

Users Who View the “Learn More” Page Are More Likely to Use Continuous Recording

Our data analysis shows that users who read the Learn More page had a significantly higher rate of using the Continuous Recording feature than those who skipped it.
Users in the Learn More group were almost 2X more likely to enable the feature than those who skipped it (15% vs 8%). This indicates that users who are informed about the value of Continuous Recording are more likely to engage with this feature.

Continuous Recording Official Launch

Promotional e-mail

Projects

english

中文

Story

challenge

Ideation

experiment

optimization

takeaway