Clean Master

Private Gallery

Private Gallery is a new feature in Clean Master. I design the overall information architecture based on the product requirements, and after the feature launched, I also increase the sign-up CVR by improving the efficiency based on the KLM-GOMS.

  • Design Strategy/ Information Architecture / Spec Guideline / Data-based Iteration

  • 5-week development on Aug 2017 

  • Collaborated with 2 Product Managers/ 1 UI Designers/ 7 Android Developers/ 3 Quality Analysts


Private Gallery is a new feature that helps the users to encrypt their pictures which contained privacy concern by preventing snoopers from viewing them in system album.


However, the product goal was to develop at a fast pace to verify the user needs. Therefore, the challenge for me was to clarify all information architecture and the product logic in a considerably short time. My responsibility was to discuss problems with PM and RD frequently and note all specs in my flowchart so as to help all stakeholders understand the architecture more clearly.


Main Flow

The main design goal of the architecture is to ensure the rational logic and the process effectiveness.

Video import process

When I designed the video import process, I fixed the stagy transition between two apps by the animation with a minimal development effort.


Create shortcut

In an attempt to help the user enter Private Gallery more efficiently, I considered several scenarios to guide user create a shortcut on the home screen.


After the Private Gallery launched 3 months, we’re happy to see that there are approximately 1 million sign-up users, and we’ve encrypted almost 50 millions photos! Moreover, the product retention is over 80%, hence, we do validate the user needs in the private photos.

Data-based Iteration

Nevertheless, according to the data statistics, we found that the conversion rate in email sign-up process was only 44%.


The main goal is to improve the email sign-up process, to help more users enter the landing page.


I analyzed the previous problem in the task was the low efficiency, which means the user has to type the email address manually during the process. According to the research of KLM-GOMS, a user types each alphabet with mobile phone cost 0.6 seconds, so if user type “”, the whole task would take 18.85 seconds to complete.

GOMS is a widely used method by usability specialists for computer system designers because it produces quantitative and qualitative predictions of how people will use a proposed system.


In order to increase the efficiency, I made a design that users can preview their email address before typing. The concept behind this is that we’re able to track the existed email logged in the phone, so user can tap it directly without typing manually. On the other hand, if we cannot track the email, the text field would show the relevant address when the user types after the “@”, then it can reduce the typing time as well.

At last, I convinced the stakeholders with my design by KLM-GOMS. With the same email “”, the shortest time with the new design only takes 7.45 seconds to complete. Even when the case that we can’t track the email, it still only takes 16.5 seconds and is shorter than the original one.

new design

new design_can't track the address


We did A/B testing first between the original and the new design. After observing the data for a week, we’re glad to see that the sign-up conversion rate increased from 44.06% to 56.97%  based on our new design. Therefore, we released all online version to the new one.


Since you made it this far,

You may be interested in this project.

------------- or -------------

Big Thanks For Visiting!

Feel free to EMAIL me to get connect