How does Didi read passengers’ reviews?

How does Didi read passengers’ reviews?

In the era of digital travel, passengers’ evaluations have become an important basis for online ride-hailing platforms to optimize services. As an industry giant, how does Didi treat passengers’ evaluations? This article will analyze Didi’s evaluation mechanism from a data perspective, and explore the operational logic behind it based on hot topics across the network in the past 10 days.

1. Core data of Didi passenger evaluation system

Didi's rating system adopts a 5-point scale, and passengers can evaluate the service attitude, vehicle condition, driving skills and other dimensions. The following is sample survey data in the past 10 days:

2. Three major mechanisms for Didi’s handling and evaluation

1.real-time feedback system: Negative reviews trigger automatic alerts, and customer service will intervene within 20 minutes. Data shows that 85% of complaints are responded to within 2 hours.

2.Driver reward and punishment system: Drivers with a score lower than 4.0 will face training assessments. Those with a score of 4.9 or above for 3 consecutive months will be awarded the title of "Gold Medal Driver" and traffic tilt.

3.AI sentiment analysis: Perform NLP processing on text evaluations to identify emotional tendencies. The recently upgraded algorithm can accurately identify 93% of implicit dissatisfactions (such as vague comments such as "It's okay").

3. Evaluation management in recent hot events

Combining hot search topics across the Internet, Didi demonstrated its evaluation management strategy in the following events:

4. The actual impact of user reviews on Didi’s operations

Data shows that passenger reviews directly drive the following improvements:

1.Function iteration: The "silent mode" function added in 2023 is derived from 23,000 related suggestions

2.Rule optimization: The cancellation of "extra-distance waiting fee" stems from the fact that 18% of negative reviews involve disputes over waiting fees.

3.Driver management: 79% of the illegal drivers who were dismissed in the year had more than three serious negative reviews.

5. Industry comparison and future trends

Compared with other travel platforms, Didi’s evaluation response speed is faster, but the level of refinement needs to be improved. T3 Travel adopts a "double-blind evaluation" mechanism (where drivers and passengers cannot see each other's ratings), while Meituan Taxi is trialling a "video evaluation" function. The industry is showing the following trends:

1. Diversification of evaluation dimensions (newly added epidemic prevention and disinfection, barrier-free services, etc.)

2. The proportion of intelligent customer service processing has increased (the current industry average is 68%)

3. In-depth integration of evaluation data and insurance services

Didi's recently launched "review correction" function allows passengers to modify submitted reviews, which reflects that the platform is building a more flexible review ecosystem. As its product manager said: "Every score after the decimal point is our improved GPS coordinates."

Through structured data analysis, it can be seen that Didi has transformed passenger reviews into the core driving force for platform evolution. In the era of experience economy, this "evaluation as a service" operating philosophy may become a new standard in the travel industry.