How 58.com Revamped Its Rental Listing Flow to Boost Success Rates

Introduction

58.com is a leading domestic classified information platform where user‑generated rental listings are a core business. To make posting easier and faster, the team launched a six‑month redesign of the rental publishing flow.

Problem Analysis

We conducted user interviews with a friend‑group of potential renters and tracked real users in a 58 rental user group, gathering over 30 actionable suggestions. Online telemetry showed that, despite a relatively high overall success rate, users frequently encountered prompts such as missing images, missing community name, or missing phone number, requiring 1‑2 attempts and an average of 7.8 minutes per posting.

The publishing page’s visual style was cluttered with redundant lines and mixed left‑center alignment.

The process spanned four pages (property info, additional info, contact info, commitment) with low fault tolerance, causing repeated interruptions.

Form fields lacked automation; users had to manually fill most information without predictive defaults.

Define Goals

Based on the problems, three upgrade dimensions were set: lower the posting threshold, reduce the time spent publishing, and improve overall publishing efficiency.

Design Implementation

Visual Noise Reduction – Cognitive Load Decrease

The new form adopts a simplified visual language: left‑aligned fields, smaller font sizes, larger hierarchy, and merged related information to lessen visual burden.

Operation‑as‑Creation – Easy Onboarding

Previously, users had to complete all required fields before publishing, often getting blocked by missing items. The new “operation‑as‑creation” approach allows a listing to be created after filling just one required field, with remaining fields completed later, dramatically lowering the entry barrier.

Global Flow Nodes – Better Process Control

We re‑structured the multi‑page form into progressive nodes that indicate publishing progress, prioritize essential fields, and provide contextual prompts, improving fault tolerance and guiding users to the next step.

Template‑Based Detail Input + Smart Fill

By leveraging existing property data, the system can infer and suggest house type, orientation, elevator availability, and rent range after the user enters community and area, auto‑filling these fields while allowing corrections. This also generates a templated property description, reducing manual entry and avoiding sensitive‑word blocks.

Summary Review

Combining qualitative user feedback with quantitative telemetry uncovered multiple pain points across the posting journey. By simplifying visual design, introducing progressive flow nodes, and applying smart form filling, the redesign reduced cognitive load, shortened publishing time, and increased successful postings. Ongoing user listening and iterative optimization remain essential for further improvements.

Thank you for reading.