 ==>> 点击下载文档 |
我们以武汉市所有高校学生和城镇常住居民为调查对象，采用分层三阶段的
抽样方式，保证了各层最终单元入样的概率均相同。预调查问卷各项检验均通过，
因此问卷结构合理。正式调查共覆盖了10所高校和5个行政区，共计发放了859
份问卷，回收有效问卷770份，总回收率为89.64%。我们对正式调查所得数据
进行了信度、效度分析，同时数据也通过了游程检验和独立性2检验，因此所
得数据随机化程度高、调查项目之间也是显著相关，所以调查结果真实可信
我们运用了描述性统计的方法分析了餐饮外卖O2O平台的使用人群的基本
特征。利用二元选择模型、计数模型，分析了影响餐饮O2O外卖平台是否使用和
使用次数的原因，运用非线性典型相关分析的方法挖掘了潜在用户不使用的原因
与用户的特征。最后，使用网络数据挖掘的方法获取微博中的信息，进行分析
通过分析得到如下结论：第一，卫生问题不是用户最不满意的问题，相反物
流问题更值得重视。第二，潜在用户中居民户群体的挖掘潜力较大，而且根据行
政区的不同表现出不同的特征。第三，年龄、收入、网购次数、网上支付工具的
使用频率和婚恋状况对是否使用餐饮O2O外卖平台有显著影响。第四，对于高校
学生，性别、生活费、网上支付工具的使用频率和订餐经历都是影响订餐次数的
重要因素；对于常住居民，年龄、收入、网上支付工具的使用频率和有无小孩是
影响订餐次数的重要因素。对此，提出以下建议：一，重点完善物流体系来维持
现有客户；二，区域差异化营销以争取潜在客户；三，根据使用人群偏好对平台
进行改善；四，根据使用者的特征对其进行精准营销
根据以上分析，餐饮O2O平台若在潜在用户的争取上采取针对性地策略后
其覆盖人群也会变广，规模也将不断扩大
关键词： 餐饮O2O外卖平台；分层三阶段抽样；二元选择模型；计数模型；
非线性典型相关分析；网络数据挖掘
餐饮O2O外卖现状分析及前景展望
-2-
餐饮O2O外卖现状分析及前景展望
-3-
Abstract
As in May 2013, America's two biggest takeaway website Grubhub and
Seamless announced a merger, the online takeaway market began to attract
people’s attention.Delivery is in boom development, and China's three
big Internet giant website baidu, tencent have dapple in the area. However,
at present our country's food industry is still in the weak link.The online
take-away mode distribution lack standardization and concern for
restaurant food safety is increasing so it the time for management mode
and ways to change and innovate.
We take all the college students in wuhan city and town residents as
investigation object, and adopts stratified three-stage sampling
probability sampling method. For the first stratify we will select the
universities and colleges firstly ,then select the dormitories Households
and finally select students. For the second stratify, we will choose the
administer zone first then choose the living quarters and finally the
household to ensure the final units on each floor into the probability
of sample are the same. We carefully design the questionnaire, so the
questionnaire structure is reasonable.Formal investigation covers a
total of 10 colleges and universities and the 5 boroughs, a total of 859
questionnaires, recycling effective questionnaire 770, the total
recovery rate is 89.64%.We adopt a formal investigation method to analyze
the reliability and validity of the data, including runs test and
chi-square test , so we can ensure the result is convincing.
We use descriptive statistical method to analyze online food
take-away platform to acquire the basic information about the using of
the crowd.Then we use binary choice model, counting model, nonlinear
canonical and correlation analysis method, to find whether people use and
the characteristics of people who prefer the form.Finally, we use the
network data mining method to get the information from weibo, to
complement the results of sampling survey.
Through the statistical analysis we get the following conclusions:
first, the potential users of households group mining potential is bigger,
and depending on the administrative region showed different
characteristics. Second, the health problem is not the most disgruntled
users, on the contrary logistics problems deserved more attention. Third,
。。。以上简介无排版格式，详细内容请下载查看
