Thunder PC 12.0

As the product declined, users increasingly described it as complex, information-heavy and overloaded with ads. Years of accumulated tools, services and commercial placements had weakened the hierarchy around downloading.

Downloading was still the main reason people opened Thunder. The reconstruction was therefore not a visual refresh or a removal of every secondary feature. We needed to decide what users should see first, what could be collapsed, and where commercial content could appear without interrupting the core task.

Core downloads, low-frequency utilities and commercial placements competed for the same attention. Expanded task details occupied most of the screen and hid the content below. Scattered commercial placements reinforced the perception of too many ads.

We compared an extremely minimal direction with a file management direction. The minimal version made downloading feel lighter but weakened cloud and downstream consumption. File management supported organization but risked turning the download product into a file browser.

We chose a controlled reconstruction centered on downloading. Task details were collapsed, low-frequency tools moved out of the primary path, and the relationship between downloads, transfers and cloud tasks was reorganized.

Commercial placements were reduced and repositioned to limit interruptions during core download tasks.

Private Space users showed stronger download activity and retention, so we strengthened its entry and onboarding to make the capability easier to understand and adopt.

In the version experiment, Private Space usage in Thunder 12 reached nearly 12 times the Thunder 11 level. Overall retention improved while core download activity stayed stable. Payer count increased 8% and payment amount increased 5%.

A multi-file download can contain dozens or even hundreds of files. The original panel often selected low-value files by default, while important content appeared deeper in the list.

Users had to open the full list, find the files they needed and repeatedly correct the default selection. The panel already offered enough control. The problem was that the system's first selection created more work.

We redesigned automatic filtering, default selection and file ordering across Windows and Android.

The new logic filters obvious low-value files, selects content that is more likely to be needed and moves selected files to the top. Manual correction remains available for special files and individual preferences.

The intended shift was from reorganizing an entire file list to checking and confirming a more reliable default result.

On Android, 19% fewer users opened the full list, while P90 time from panel display to download click fell 7.26%. On Windows, file selection corrections fell 12%.