The data source of this report is the take-away interview questions of Meituan. Excel is used for data cleaning, and Power BI is used for visual analysis.

The analysis is as follows:

The analysis idea mainly consists of the following four points:

1. Analysis of bad label: what is the proportion of different labels;

2.? Site analysis: analyze the performance of different sites and existing problems;

3.? Rider analysis: analyze the number of bad reviews of riders, delivery time indicators and related factors with the site;

4.? Merchant analysis: analyze the problem of bad reviews caused by merchants.

1.? Analysis of bad review labels

2. Site analysis

? The problem of delivery timeout accounts for a serious proportion in every site. Site c performed worst, followed by site a and site B.

the total time consumption of sites b, c, e and f is too long, which may easily lead to overtime delivery of orders.

although site a basically takes a short time, it has the second largest number of bad reviews. Combined with the proportion of bad reviews, service attitude and other problems are more serious.

3.? Analysis of riders

Overtime delivery and bad attitude are the common problems of most riders.

riders are more likely to get bad reviews if they eat less/spill meals, deliver them earlier and have uneven appearance.

the single dimension of the rider's delivery time is not the decisive factor for customers to give bad reviews.

there are problems in the management of stations a and c, and the same rider gets more bad reviews at these two stations than at other stations.

4.? Merchant analysis

? It's not all about the rider's service that each rider gets bad reviews, and the restaurant business also has certain responsibilities.

analysis conclusion and improvement scheme.