Mass data analysis is the core of electric Shang

At present, there are many enterprises to expand the electric business, but the success rate is quite low, not more than 10%. Many of the enterprises doing electric business, there are large scale, brands, have the strength of large enterprises, they do not lack the funds to do electricity, but still failed to get rid of the fate of failure. Here need to recognize the fact that the success of E-commerce is not entirely a matter of capital, more important is good at using methods.

The author led by a company assisted by a retail enterprise as an example, the project in the third party marketing costs the lowest situation, but to create the fastest growth of performance (compared with China's retail hundred) E-commerce platform. In this case, the enterprise is one of the Chinese retail industry hundred strong, in China's retail e-commerce lack of success stories and successful path of the background, the enterprise in the beginning of E-commerce is more worried and hesitant. And in our assistance, created in the third party marketing costs of not more than 300,000 of the case, in the last August after the official launch of the 6 months, to achieve cumulative sales of over 30 million results. This achievement, the core reason is that we capture and analysis of massive data, from the beginning of planning through a large number of data analysis to determine the selection, pricing and related marketing activities.

What can be sold through massive data analysis and screening

Retail involved in the product category and many brands, the selection of products facing the very complex. In deciding to do the retail electricity business, we have to determine in advance what the Internet to sell, what products are consumers need, what products will be hot. In the past, the stores that reacted to this demand quickly gained an advantage. But now, good retailers are good at using huge amounts of data to gain competitive advantage, predict trends, and prepare for future demand. Want to get the month's selling, you have to start a few months ago the site selection.

In the product selection on the use of PCs product strategy, through a large number of data collection, the establishment of a correct model to determine the product. First, from commodity attributes, commodity latitude, market sales, industry in the future and leading enterprises, such as the five levels of screening, from hundreds of product categories, through the operation, screening, validation process, to obtain the best of several product categories; After the whole network information retrieval, collection of major network platform related categories of commodity information, sales of items, Sales, turnover volume and other data; then the above attributes are different, the parameters of the massive data, adjusted and simulated to the same environment, so that the same as possible real reaction to the quality and disadvantage of the same condition, to see whether the product is suitable for online sales; and then through the socioeconomic data parameters for further revision, so that more consistent with the macro And finally through the third party research institutions data sharing and joint research, again to amend, thus more in line with the market environment. Through this process, determine the customer group and business positioning, and finally determine which products to sell.

Determine how much money to sell through massive data analysis

After the product is determined, also need the reasonable product price. The price of a product is not based entirely on the cost price, or the price follows the strategy. The price of similar products in the face of the situation is: prices set high no one, set low and earn no money. This is a contradictory proposition, how to find a balance between the relationship between the use of the price strategy, which also need to rely on massive data analysis.

In the pricing of products, the product price model is established by collecting the data of corresponding products and combining with other data, including regional economic data, population characteristics, network situation, commodity information, sales hotspot and so on. According to this to determine the corresponding product price bands and price lines, to establish a reasonable price of products. On this basis, determine a certain range of pricing, and see whether this pricing will be diverted from other pricing products sales, thus optimizing the pricing strategy. At the same time, according to marketing activities and seasonal changes in demand for products using dynamic prices, through such a price strategy, in the price increase a certain proportion of time, but also to maintain sales, reduce inventory. On the other hand, it can adjust the reasonable proportion of the high profit margin goods while keeping the consumers to maintain the market share.

Decide what marketing activities to do through massive data analysis

In order to attract customers, promote sales, various types of promotional activities on the Internet, such as: How much discount, buy how much to send, how much less how much, send red envelopes fu bag, send coupons, seconds kill, auction, etc., play more, but the same phenomenon is very serious activities. Admittedly, these activities are indeed attracting consumers, to the site attracted the flow, but in the realization of profitability, it is not necessarily satisfactory. In this case, it is clearly not feasible to follow blindly and do activities for activities. And how to combine the characteristics of their own website, to carry out attractive, easy to carry out, high credibility of the online activities will also be based on the massive data analysis.

The data model between the activity type and the activity effect is established by capturing the mass data associated with the activity. According to the large number of customer transactions captured, demographic data, shopping patterns, regional hotspots, online transaction data, physical store sales data, to determine the ROI of marketing activities (ROI), predict the use of how to better marketing tools, determine the most appropriate marketing methods. And for the case of the entity store enterprises, using a convenient implementation of the line, the combination of offline activities. Online and offline disclosure of activity information, attract a large number of online registration, and through the activities of the Association, to promote the registered members of the purchase behavior and interactive participation, and can participate in offline interactive activities. Activities to promote the registered members of the steady growth, and maintain a stable sales. In the event of stimulation, plus its own selection and price advantage basis, in the open sales, the establishment of consumer trust, and maintain the site sales, to achieve the best return on investment.

Massive data analysis ultimately needs to be combined with psychological use

In general, a city's data analysis needs to reach at least 200 million. Our PCs product strategy, for example, is to select only a few of the hundreds of kinds of products, and the amount of data can be imagined for the price of thousands of goods and products. To obtain more accurate analysis results, we need data capture experts, information analysis experts and rule recognition experts and other cooperative efforts. The use of large numbers of good enterprises, can win the market. We are familiar with a number of brands or retail enterprises, are successful use of large numbers of data analysis enterprises. Amazon, for example, based on a massive data analysis of precision marketing, Wal-Mart analysis of the massive data on social networking sites to reveal consumers ' preference for demand. The success of our case is also due to the emphasis on the analysis of massive data from the beginning, from the selection, pricing and marketing activities.

However, it is necessary to point out that mass data analysis is not a simple summary of historical experience, because any business activities can not be separated from the "people", need to combine the psychological and behavioral studies to accurately predict the results. Through the consumption psychology to study the individual psychological activity law and the personality psychological characteristic, through the social Behavior Study to the group consumer socialization behavior Research, unifies these studies to the data analysis result, and applies to the electronic commerce operation practice, guides the consumer the consumer behavior, finally realizes the electric trader to win.

It can be said that all e-commerce needs to do a large number of data analysis, enterprises need to pay attention to data and related analysis, with analytical conclusions to guide the development of the electrical business.

Author Introduction:

Jin Xin, China's leading E-commerce leader, national experts. Zhi Xin and the public (e-commerce direction), Zhi Xin Yu Jie, the founder of cloud mud technology. E-mail: service@ec-serve.cn, Welcome to Exchange.