Entrepreneurship is a little risky. Read the data and do a business analysis

1、 cause

After the new year , As a program is also very old , So I want to do a sideline , But blind entrepreneurship is not enough , Mr. Lu Xun said ： Direction over effort , So choosing the right direction is very important , The first step of data research , Don't act blindly , A waste of time , Waste energy , The main waste of my hard-earned money .

I usually order takeout during working hours , So I want to make takeout , Because I am Xiaobai , So I want to find a brand takeout store to join , But most of them know that franchise stores are relatively stupid , So be cautious , Do research .

2、 Data crawling

The platforms selected for data crawling are two big takeout platforms , Here, choose one of them for analysis. I want to see the take away order volume of the business district , Do a survey , The data content crawled is the category of takeout , A single quantity , And the name , Location , These basic data , And then make a summary , Analyze the data in the business district .

3、 Crawling step

1、 Confirm the URL you crawled

The data source of crawling is a app, The local solutions are nocturnal simulator and Charles, Installed the environment and did a few things , I didn't figure it out , This plan was abandoned

stay xxx After a search on our official website, we found h5 The interface of , It seems that this interface is not public , After searching for a long time, I finally found the entrance

https://h5.xxx.me/

After twists and turns, I finally found , But after logging in, you still have to jump to the official website , So enter this URL again  

be based on csdn The rules of ,xxx Please replace

  The final crawling data interface is ：

https://h5.xxx.me/restapi/shopping/v3/restaurants?latitude=31.296829&longitude=120.736135&offset=16&limit=8&extras[]=activities&extras[]=tags&extra_filters=home&terminal=h5

 2, Analyze the website

The data request method is get request , So all the parameters are url in

latitude=31.296829

&longitude=120.736135

&offset=16

&limit=8

&extras[]=activities

&extras[]=tags

&extra_filters=home

&terminal=h5

The first two data are longitude and latitude , That is, the offset of positioning information , This data determines the data pulled , This is where I want to locate .

offset I don't know what this offset means , I guess it may be the number , Is not important

limit I guess how many data per page

terminal Is the type of terminal , Basically

3、 Analyze the returned data

{
has_next: true,
items: - [
- {
restaurant: + {... }
},
- {
restaurant: + {... }
},
- {
restaurant: + {... }
},
- {
restaurant: + {... }
},
- {
restaurant: + {... }
},
- {
restaurant: + {... }
},
- {
restaurant: + {... }
},
- {
restaurant: + {... }
}
],
meta: - {
rankId: "",
rankType: - {
505517688: "33"
}
}
}

The most external data format is like this , What we want is restaurant This layer of data , It is also important to analyze  

restaurant: - {
act_tag: 0,
activities:[],
address: null,
authentic_id: 502101541,
average_cost: null,
baidu_id: null,
bidding: null,
brand_id: 710858,
business_info: "{"pickup_scheme":"https://tb.xxx.me/wow/a/act/eleme/dailygroup/682/wupr?wh_pid=daily-186737&id=E14670700902593244","ad_info":{"isAd":"false"},"recent_order_num_display":" On sale 1244"}",
closing_count_down: 14234,
delivery_fee_discount: 0,
delivery_mode: + {... },
description: null,
distance: 2804,
favor_time: null,
favored: false,
flavors: + [... ],
float_delivery_fee: null,
float_minimum_order_amount: 20,
folding_restaurant_brand: null,
folding_restaurants:[],
has_story: false,
id: "E14670700902593244",
image_path: "https://img.alicdn.com/imgextra/i2/2212739556234/O1CN01jzFa3R1vvDoDnsojl_!!2212739556234-0-koubei.jpg",
is_new: false,
is_premium: true,
is_star: false,
is_stock_empty: 0,
is_valid: null,
latitude: null,
longitude: null,
max_applied_quantity_per_order: -1,
name: " Fruit cutter （ East Lake CBD shop ）",
next_business_time: " Tomorrow,  9:30",
only_use_poi: null,
opening_hours: - [
"9:30/0:10"
],
order_lead_time: 42,
out_of_range: false,
phone: null,
piecewise_agent_fee: + {... },
platform: 0,
posters:[],
promotion_info: null,
rating: 4.7,
rating_count: null,
recent_order_num: 1244,
recommend: + {... },
recommend_reasons: + [... ],
regular_customer_count: 0,
restaurant_info: null,
scheme: "https://h5.xxx.me/newretail/p/shop/?store_id=546110047&geolat=31.296829&geolng=120.736135&o2o_extra_param=%7B%22rank_id%22%3A%22%22%7D",
status: 1,
support_tags: + [... ],
supports:[],
target_tag_path: "35a1bb9025ab98c28112d82f83f73d7ejpeg",
theme: null,
type: 1
}

The data seems easy to understand , The data we need here is

 name The name of the take out shop

business_info.recent_order_num_display  Sales figures

activities Store activity data

opening_hours It's the business hours of the store

support_tags It is the classification of stores

4、 Crawl data

I haven't written all the code for crawling data , So the above analysis

The core of crawling data is pretending to be a normal request , We have seen the data web Interface , Basically nothing

In general, that is :

cookie requirement

js encryption

token Request and so on

Because I don't use it in batches , So it can be web After the client logs in , Use data directly , Disguised as a web That's all right. , At its worst, it can be used directly seleunim To climb , So it's not a big problem

5、 summary

        Looking at the data, it seems that some brands have good sales , But I can't see the opening time of the store , So there is no way to judge whether it is a buying behavior , But the average sales volume is not high , A little polarized , The situation of taking out is grim , Earn hard money , The platform also draws from the turnover 15% , Remove the franchise fee , Shop rentals , There are also various equipment costs , It's really a lot of money . In the end, all the money was earned by the platform , As a business, there is basically no profit . therefore Take out is not recommended , Feel stifled .