How a Design Sprint Transformed NetEase Yidun’s Image Filtering Experience

Design sprint methods, popular in the Bay Area, promise rapid solutions to tough product problems. The article shares the first practical attempt of applying this Google Ventures‑originated 2‑5 day process to NetEase Yidun’s image‑online experience.

Design sprint combines design thinking and agile development to accelerate product development, focusing on validating user needs and fostering collaborative innovation.

Six Phases of the Design Sprint

Understand – interview stakeholders, observe users, map interests, and analyze competitors to grasp user, business, and technical requirements.

Define – pinpoint the core opportunity and set clear goals.

Diverge – brainstorm a wide range of ideas without judgment.

Decide – vote on sketches and select the most promising concept.

Prototype – build a low‑fidelity prototype for user testing.

Validate – test the prototype with target users and gather feedback.

B2B vs. B2C Considerations

Unlike consumer‑focused (B2C) products, B2B solutions serve rational enterprises that demand efficiency, professionalism, and clear ROI, requiring adjustments to the standard sprint steps.

Case Study: NetEase Yidun

NetEase Yidun, a one‑stop B2B security service, faced low efficiency and poor user experience in its image‑online detection module. The team customized the six sprint steps to address these issues.

1. Understand

Through business goal interviews, technical feasibility checks, user interviews, field observations, stakeholder maps, and competitor analysis, the team defined the problem: improve the usability of the image‑online experience.

Who cares about the online experience?

What constitutes a satisfactory online experience?

What are the goals of this module?

What information do customers expect?

The target users were identified as UGC products needing anti‑spam image filtering, with a core need to test the precision of ad‑image detection.

2. Define

A user‑journey map highlighted that users focus on the detection process and results, emphasizing efficiency and professionalism. The sprint goal became delivering a simple, engaging, high‑satisfaction experience that solves efficiency and professionalism challenges.

3. Diverge (Brainstorm)

Team members generated ideas such as visual indicators, scoring systems, random image libraries, hover‑preview galleries, and feedback mechanisms. Sketches were created for each concept.

4. Decide

Votes were cast on the sketches; the hover‑preview gallery (Solution 2) received the most support and was chosen as the final solution.

5. Low‑Fidelity Prototype

The prototype integrated left‑side navigation, sample explanations, multiple upload methods (local file, URL, drag‑and‑drop), detection status, and result details with reasons and percentages.

6. Validate

Five target users were recruited. After a brief warm‑up, participants performed the image‑filtering task while the team recorded observations and scored each feature. All participants completed the tasks smoothly, reported high satisfaction, and confirmed the solution’s effectiveness.

Conclusion

Design sprint proved to be a highly efficient research and validation method for B2B products, enabling rapid problem definition, idea generation, decision‑making, prototyping, and user testing. By adopting this disciplined approach, teams can focus on real user problems, iterate quickly, and improve overall workflow.