How 58.com Boosted Used‑Car Phone Conversions by 95% Through User‑Mind‑Driven Design

01 Business Background and Design Planning

58.com’s second‑hand car service is shifting from a traditional product‑guidance model to a personalized recommendation model based on user profiles. The goal is to improve car‑finding efficiency and increase phone conversions.

Designers, as a bridge between business and users, planned two main directions:

Focus on the core metric “phone conversion increase” by capturing user mind differences at each core link and finding experience breakthrough points.

Balance business and experience metrics to explore a conversion model suitable for large‑ticket, low‑frequency items like used cars.

02 User‑Mind‑Driven Chain Observation

The business metric is “phone conversion.” Users move through core touchpoints: Home (search → category → list), vehicle detail, micro‑chat, and store pages. Each touchpoint carries specific scenario value, and their collaboration ultimately enables a phone connection.

03 Designing from Business Metrics

The annual target is a N% increase in phone conversions. This target was broken down across five core touchpoints (search, category, list, micro‑chat, detail), aiming for roughly N/5 % improvement per touchpoint.

04 Enhancing Touchpoints Based on User Mind Differences

Home Page

Current state: users face a “filter first” requirement, raising the entry barrier, and the waist‑area recommendation has low exposure.

Solution: segment users into shallow and deep visitors; provide richer, more visual content for shallow users and recommendation cues (e.g., how many people chose) for deep users.

List Filter

Current issues: high usage threshold for the filter and excessive dimensions increase operation cost.

Solution: introduce a “quick filter” card displaying the four most frequent dimensions (price, type, brand, series) with one‑click access and supplemental recommendation info.

List Page

Problems: inconsistent tag styles, scattered information, and uneven item heights.

Solution: consolidate tags, prioritize “real‑car” tags, place them under images, and streamline right‑side fields to improve information density.

Detail Page

Current gaps: over‑emphasis on protection info, weak car‑condition display, fragmented information flow.

Solution: define four mind‑decision nodes—Understand, Trust, Recognize, Convert—and reorganize content to match these stages, adding concise car specs, trust‑building details, and conversion prompts.

05 Results and Conclusion

After redesign, each module exceeded its phone‑conversion goal, with the overall achievement reaching 95.5 % above the target. By capturing user‑mind differences and upgrading each chain node, the product significantly improved conversion efficiency and helped the business meet its annual objectives.