1th Store full cycle nuggets: Big data into operation Core driving force

At the moment when it comes to http://www.aliyun.com/zixun/aggregation/8213.html "> Big Data Applications, the first thing people think about is marketing push, which seems to be the main value of big data." But the practice of Shop No. 1th shows that large data can be the core driving force of operation.
The 1th store website as the enterprise interacts with the consumer portal, each day carries tens of millions of merchandise clicks, browses and buys, gathers into the massive data. For shop number 1th, this is the basis for improving operations.

1th, vice president of product design Wang Xinlei said: "Consumers come in is how to use, how to find goods, how to pay, the whole process, in the use of large data analysis, and then make corresponding improvements." "Its application to large data runs through the introduction of customer traffic, leads customers to purchase, to enhance the buyer loyalty of the whole customer life cycle."


for electric business enterprises, how to introduce traffic from the Internet to their own site, is the starting point of operations. The first is where the customer comes from, the key is three dimensions: one, the customer from which channel; second, the customer from which area; third, customers from which user groups, new users or old users. The analysis of these three dimensions directly determines the input of the follow-up drainage resources of store 1th, which are rooted in the large data analysis of customer behavior in shop 1th.


in analyzing which channel the customer is from, through the site to collect a large number of customer traces, store 1th can find more traffic and need to strengthen the channel: micro-bo, forum, or portal, so that constantly adjust the marketing, such as to find which channel can put more ads, which channels have potential, has not been fully tapped. In analyzing the areas from which customers come, through large data analysis on the source traces of customers on the website, store 1th can find those areas where sales grow fast and grow slowly, and adjust the marketing cost of different regional markets accordingly; If the customer is a new user or an old user, if the website browses and buys the data more from the old user, Companies can reduce market costs accordingly.


Large Data marketing push is also a very important source of traffic, No. 1th store in addition to a large data method for consumer classification modeling, but also created a shopping list mode. There is a shopping list next to the search box at store 1th. The goods that consumers have purchased at shop 1th are displayed on the shopping list, and consumers can add them separately. For consumers, this is easy to buy, and for businesses, shopping lists are a big source of data that reflects consumer preferences. Through the shopping list of data, shop number 1th in accordance with the consumer purchase cycle, the consumer marketing recommendations. For example, a customer looked at the goods, did not buy, but added a shopping list, when the merchandise discounts, store 1th will be in time to push customers.


When the customer enters the No. 1th store, it enters the buying stage for the customer. This stage, how to enhance the purchase amount of each customer, and in this process, the realization of goods and the optimal allocation of resources, is the key to operations. The big data once again became the hand of shop number 1th. First of all, 1th Shop's website improvement, including pictures, web design, completely to customer clicks and browsing, and other behavioral traces of large data analysis to rely on. Moreover, in the process of interacting with consumers, store 1th also used large data. Like some shopping malls shoppers, consumers browsing the Web site merchandise process, shop number 1th will give consumers some tips recommended, according to consumers before the browsing and purchase behavior, the 1th store system can determine what consumers may like the goods, give the corresponding tips. Again, according to the consumer is to search for goods, or browse merchandise, shop 1th can initially determine whether he is a purposeful, limited time buyers, or time is abundant, purposeful buyers, for the former will directly recommend the goods, for the latter, then continue to stimulate their buying behavior.


customers to purchase goods, into the follow-up logistics and distribution services. At this stage, how to achieve the best supply chain efficiency, reduce warehousing and distribution costs, improve delivery speed, e-commerce business is the lifeblood of the operation. How to achieve more efficient picking, is the key to the impact of logistics efficiency. Shop number 1th creates an efficient way to pick goods--pick up the goods. Customers tend to buy several items at a time, and if an order is picked up once, the pickers may repeatedly pass through the same area and waste a lot of time. So shop number 1th stores close orders for several items in the area so that the pickers can pick up all the items associated with an order in a single area. In the allocation of goods, how to let the pickers take less road, you need to rely on large data analysis. First of all, store 1th use large data analysis to find the highest degree of overlap in the order group, such as consumers buy the same category. Second, when placing goods, consumers often buy together, high degree of aggregation of goods together, such as Cola and French fries. This kind of logistics arrangement based on large data greatly improves the efficiency of picking, at present, there are 16.7 items on average in shop 1th, and employees pick a single item in less than 80 seconds.


distribution, how to provide the corresponding service options, how to charge, but also based on the analysis of large data on the basis of 1th stores the latest distribution services "Four days a day, six to send", can allow consumers to specify a special distribution time. And whether consumers like such distribution services, will not use, completely rely on consumer traces of large data analysis. Shop 1th will look at the number of consumers who click on this option, how many people use this service, the percentage of clicks and the last actually used. If the click is not much, on behalf of the distribution service is not attractive, if the number of clicks, the actual use of not much, it may represent the cost of this service a little higher, need to consider adjustment costs.


For businesses, the end of consumer buying does not mean the end, but also the desire to turn consumers into their loyal customers. At this stage, shop number 1th is also fully releasing the power of large data.


store 1th found that after buying three or four orders, consumer loyalty became quite high. To do this, it needs to constantly push customers across the threshold, but first find out which customers are most likely. Store 1th with large data analysis to screen out such consumers, and accordingly through a number of concessions and points to buy the desire to stimulate the purchase of these purchases, to promote their purchase of the Third Order, the fourth order.


1th stores also rely on large data mining and analysis to reduce customer churn, to those who may lose customers, through a number of targeted wake-up and retention action to stimulate, customers birthday, will wish their birthday happy, or send some promotional information, to arouse the customer's perception of the site. Timing is also dependent on the customer purchase cycle of large data analysis, time is too early, may do no work, wake-up time too late, and may not be late.


1th Why can we achieve a deep combination of large data and operations? Through research, we find that this is inseparable from the centralized and transparent system of store No. 1th. It will come from the front end of the consumer, and the large data stream from the back end of the commodity is collected and shared into various departments in real time, driving each department to operate. This is already envisaged in store 1th's initial information system planning. As chairman Gege said: "Initially, I think we have to do a centralized information system, rather than piecemeal, so that each module can share data and achieve better synergy." ”