How 58.com Revamped Local Service Listings to Boost User Experience and Revenue

Introduction

Compared with vertical business lines such as real estate and recruitment, 58 local services cover almost every aspect of daily life. Most services lack standardized processes or closed loops, making the presentation of local service information extremely complex and difficult to fit a unified design, which raises experience and maintenance costs.

This situation cannot be solved in the short term. On one hand, a single service provider’s reach is limited by geography, and providers operate in isolation, making it hard to form large cross‑regional providers. In a traffic‑driven internet economy, providers must rely on platforms like 58.com for traffic dividends and revenue. On the other hand, services are hard to quantify and standardize, so designers find it difficult to deeply optimize them. Consequently, design remains driven by connection effectiveness.

Unlike vertical services that give designers more room for value‑added features, every experience optimization in a connection‑oriented scenario must carefully consider its impact on connections. Designers therefore need to balance user experience and commercial revenue from a macro perspective.

Project Background

The service list and detail pages are the main decision‑making scenes; their design quality directly affects connection outcomes. An initial full‑process walkthrough uncovered 38 issues, most concentrated in the list and detail pages (see Figure 1). The focus of this article is on list‑page optimization.

The old list page suffered from saturated information layout, weak sense of quality, limited configurability, and an unusable filter that presented mostly irrelevant information, lowering decision efficiency. Underlying reasons include lack of long‑term business planning, low merchant participation, and absence of a user‑centric view. The old list could no longer meet user experience, commercial revenue, or business needs, making redesign urgent.

Design Perspective Shift

After learning that the local service list had undergone multiple cautious revisions, the author realized that merely tweaking framework or presentation would not achieve both user experience and commercial success. Instead, the focus shifted to how information matches user needs, using this as a breakthrough point.

First Thought Line

User needs are inseparable from the user’s context. Matching information based on user behavior, space, and time can be observed everywhere:

Mining User Behavior – Capture user actions, label them, and match them with labeled services to perform an initial filter, dramatically narrowing information and facilitating efficient decision‑making.

Based on User Space – Users exist within a physical or abstract space (e.g., location, weather, network signal). Location‑based services such as weather alerts illustrate this.

Based on User Time – Certain time slots (e.g., 8‑10 am, 12‑2 pm, 6‑8 pm, 8‑10 pm) see higher mobile usage, making them optimal moments for information push.

Second Thought Line

Narrowing the information scope is the first step; the final step is to decompose user needs into atomic elements and match each with information. For example, a phone‑purchase need becomes a set of attributes (brand, price, configuration). While filters address this, the list itself can also reflect these atomic elements, improving connection intent and revenue.

Feasibility Exploration

Each local‑service post functions like a merchant’s storefront, containing almost all service details. This creates decision friction for users but also provides a basis for atomizing post information and matching it efficiently with user behavior.

The new 58 homepage demonstrated a “thousands of people, thousands of faces” matching strategy, delivering personalized content and positive results. Extending this to local services requires breaking posts into minimal information units (user‑need elements) and applying a matching formula: Match% = (Provided Elements / User Need Elements) × 100%. The design and product teams agreed to test this approach.

Design Direction

Observing the old post’s fixed information, the author noted that if a user’s decision dimensions are A, B, C but the post only shows A, the match rate drops to 33 %, causing lost commercial opportunities. The proposed solution:

Pre‑filter users with prior behavior (search, click, collect) to narrow merchant scope; users without prior behavior see popular categories.

Modularize post information, enabling flexible presentation and better extensibility.

Defining Information Scope

All possible post fields were collected and categorized. Core visible elements include image, title, price, and rating. Rating was omitted because most merchants cluster around 4.6 stars, and displaying low scores could harm revenue. Good‑review rate was also unsuitable due to low distribution.

Design Output

The new list adopts modular, configurable information blocks that adapt to user behavior. Layout emphasizes extensibility for future modules. Key elements:

Header image based on user behavior.

Activity badge placeholder.

Title free for merchants; when matching search terms, prepend keywords.

Tags configurable; first tag shows matched keyword when applicable.

Price displayed according to user behavior priority.

Filter Optimization

The old filter suffered from poor usability, excessive depth, and non‑exhaustive, non‑independent options. Data showed users typically use only 2‑3 levels. Optimizations include:

Simplify filter dimensions and depth.

Refine scenarios to lower decision cost.

Strengthen filter guidance to indirectly increase connections.

After redesign, the filter was transformed into card‑style scenes, guiding users more effectively.

Overall Adjustments

Other adjustments include integrating self‑operated brands, recommendations, and operation‑zone entries. Important information is placed along a visual vertical line to improve browsing efficiency. High‑impact categories adopt a large‑image mode for easier selection. The final new list design is shown below.

Results

After launch, core metrics such as call/UV and cash/UV improved significantly. The new list has been rolled out to the top 20 categories. Issue‑resolution rate reached 68.4 %, addressing most pain points. The walkthrough mode has become a continuous improvement mechanism.

Conclusion

Balancing user experience with commercial revenue requires aligning design goals with business characteristics. By constructing a new information‑matching and presentation mechanism from the ground up, the redesign elevated both experience and revenue, turning a superficial “layer‑level optimization → connection drop → half‑hearted fix” cycle into a virtuous “deep optimization → efficient connection → ongoing improvement” loop.

Acknowledgments

Thanks to Huang Shan, Song Jie, Li Yuan, Zhao Jing, and all participants for their collective effort, and to readers for their time and feedback.