How does big data combine with retailing?

This era has not been a simple digital media era, some business giants have quietly used "big data" technology for many years, with large data-driven marketing, drive cost control, drive product and service innovation, drive management and decision-making innovation for many years.

Large data contains a variety of information on the operation of the enterprise, if they can be timely and effective collation and analysis, it can be a good effective way to help enterprises to carry out business decision-making, to bring the enterprise to obtain huge value-added value benefits.

This issue is the most closely linked with big data in the retail industry, and small partners to share the big data on how to truly combine with the retail industry.

The commercial value of "Big Data"

1. Segment Customer Segments

"Big Data" can be divided into customer groups, and then tailored to each group to take unique action. Targeting specific customer groups for marketing and service is the pursuit of the business has been. Cloud storage of massive data and "Big data" analysis technology makes real time and extreme segmentation of consumers is highly cost-efficient possible.

2. Simulated reality

Use "Big data" to simulate reality, explore new needs and increase the return on investment. Now more and more products are equipped with sensors, the popularity of automobiles and smartphones makes the collection of data can be explosive growth. Social networks such as blogs, Twitter, Facebook and Weibo are also generating huge amounts of data.

Cloud computing and "big data" analytics make it possible for businesses to store and analyze data in real time, along with transactional behavior, in a cost-efficient manner. Transactional processes, product use, and human behavior can be computerized. "Big data" technology can integrate these data for data mining, so that in some cases, through model simulation to determine the different variables (such as the different areas of different promotional programs) in the case of what the highest return.

3. Increase return on investment

Improve the "Big data" results in the relevant departments to share, improve the entire chain of management and industrial chain input rate of return. Large-data-capable departments can share "big data" and "big data" with weaker departments through cloud computing, the Internet and internal search engines to help them create business value with "Big Data".

4. Data storage space for lease

Enterprises and individuals have a large amount of information storage needs, only the data properly stored, it is possible to further explore its potential value. Specifically, this business model can be subdivided into two categories for personal file storage and enterprise users. Mainly through Easy-to-use APIs, users can easily put a variety of data objects in the cloud, and then like the use of water, electricity, according to the amount of charge. At present, many companies have launched the corresponding services, such as Amazon, NetEase, Nokia and so on. Operators have also launched corresponding services, such as the China Mobile cloud business.

5. Manage Customer relationships

Customer management applications are based on customer attributes (including natural attributes and behavioral attributes), from different angles of in-depth analysis of customers, understand customers, so as to increase new customers, improve customer loyalty, reduce customer churn rate, improve customer consumption. For small and medium-sized customers, specialized CRM is obviously large and expensive. Many small and medium-sized businesses will fly letter as a primary CRM to use. For example, the old customers added to the letter group, in the group of friends to release new product announcements, special sales notice, the completion of pre-sales after-sales service.

6. Personalized and accurate recommendation

Within the operator, it is common to recommend all types of business or applications based on user preferences, for example, the application store software recommendations, IPTV video program recommendations, and through the Association algorithm, text Digest extraction, emotional analysis and other intelligent analysis algorithm, can be extended to commercial services, the use of data mining technology to help customers precision marketing, Future profits can be derived from the division of value-added parts of the customer.

In the case of everyday "spam messages", information is not all "rubbish" because the person received does not need to be treated as rubbish. After analyzing the user behavior data, we can send the needed information to the people who need it, so "junk text message" becomes valuable information. At McDonald's in Japan, users download coupons on their phones and go to restaurants to pay for them with carrier DoCoMo's mobile wallet. Operators and McDonald's collect relevant consumer information, such as what frequently buy hamburgers, go to which shop consumption, consumption frequency, and then accurately push coupons to users.

7. Data Search

Data search is not a new application, with the advent of the "Big Data" era, real-time, full range of search needs are becoming more and more intense. We need to be able to search for data such as social networks, user behavior, and so on. Its business application value is to link real-time data processing with analysis and advertising, that is, real-time advertising business and the application of mobile advertising social services.

The operator's information on online behavior makes the data obtained "more comprehensive dimension" and more commercial value. Typical applications such as China Mobile's "Pangu search".

Second, "Big Data" and the combination of retail industry application

As for the use of data, many entity retailers have also expressed great importance to them, and they have made various forecasts and analyses of the data accumulated by the enterprises. However, for the specific sales business, there are often the ideal and the reality of the tangle, the market recently a well-known clothing retail enterprises on the one hand in the promotion of profit listing at the same time, on the other hand exposed to nearly 1 billion yuan inventory. Many retail enterprises in the country know the benefits of "big data" applications, but once they combine the application of "big data" into their business operations, there will be a very large problem with the current operation, so that many enterprises are very cautious attitude.

1. Combining retail strategy with "big data" technology

The biggest value of "big data" in retail business is to combine the "big data" technology in retail strategy, and to make the best of the retail strategy to ensure the realization of the sales plan. "Big data" pay attention to four "V": First, the volume of data is large (Volume), the second is the complexity of data types (produced), more involved in a variety of structural and non-structural; the third is low value density, which is relative to volume. Four is the data update and processing speed (velocity).

According to these characteristics, it is necessary to make the corresponding strategy response at the same time when the business data is generated, which will win more time and market strategy adjustment space for the enterprise. This is similar to the Big River flood peak warning, upstream peak of what the situation, downstream to do what kind of response. Data to this level, the direct business value, which is not the same sales ratio, chain, sales plan than the data can guide the business value can be compared. For example, a physical retailer involved in the online business is often prepared with 3 sets of contingency strategies during the 15-minute promotional period of a group of goods to ensure that the goods are sold as planned.

There are many cases of data and marketing in the physical business world. An earlier version is the data relationship between Wal-Mart beer and diapers in America. It turns out that American women take care of their children at home, so they charge their husbands to buy diapers for their children on their way home from work, while the husband buys the diapers and buys the beer he likes.

When analysts learned that there was a positive correlation between beer and diaper sales and analyzed them further, they found the buying situation and put together the two categories of goods. The discovery brought new sales portfolios to businesses. Of course, even if more retail chain enterprises know this story, and rarely from the usual sales can find such a combination, even if it is far-fetched.

Therefore, the retail strategy design is the retail industry "big data" the most valuable place, is also "big data" can provide directly support business.

2. Retail enterprises should maintain a correct attitude towards "large data"

The leader of enterprise should pay attention to the development of "big data" first, focus on the data center of the Enterprise, the collection of customer data as the first goal of enterprise marketing operations; Secondly, the training of internal personnel and the establishment of software and hardware mechanisms to collect data; Thirdly, to determine which data are to be collected based on business requirements; Confirm how to achieve the first three objectives of the infrastructure project on the basis of existing data or the future direction of the enterprise.

In these it foundation work needs the enterprise to have the real investment and the construction standard Information team, as the biggest part of China's business-small and medium-sized micro-retailing enterprises seem impossible and not enough ability to face such a change.

Due to the accumulation of their business and profits, large and medium-sized retailers have been able to take on the cost of such a demand trend. Small and medium-sized micro-enterprises are still in the process of rapid development, if also as large and medium-sized enterprises to carry out all aspects of investment, will soon be a new type of it tools to drag down or hit hard.

But this does not mean that small and medium-sized retail enterprises do not have the opportunity, in fact, the development of it for all enterprises to provide an equal choice, the broad application of cloud computing is a temporary gift for such a change.

As small and medium-sized retail enterprises, there is no need to consider their own building a "large data" of IT systems, they are not suitable for energy, cost, ability, so such enterprises can outsource the enterprise IT construction to the appropriate service providers, the enterprise itself all the energy can be put into the development of the business circle.

At present, some IT software development operators have also launched a cloud service base platform for traditional retail enterprises, to provide small and medium-sized micro-business enterprises with the same basic environment and system architecture for large enterprises and super large enterprise, small enterprises need to clearly plan their own goals and appropriate steps, using the cloud platform to pay on demand, Large initial inputs and unpredictable operating costs are not necessary.

The application of "big Data" in the actual combat of retail enterprises

1.Target

The first story about "Big data" happened at Target, the second-largest supermarket in the United States. Pregnant women are a very high customer base for retailers. But they usually go to specialized maternity stores instead of buying pregnancy supplies at Target. When people mention target, they tend to think of daily necessities such as cleaning supplies, socks and toilet paper, but ignore everything the target has for pregnant women. To this end, target marketers have turned to target's customer data analysis Department to establish a model that will be identified during the 2nd gestation period for pregnant women. In the United States, birth records are public, and when the child is born, the mother of the newborn will be surrounded by a deluge of product promotions and must be in the 2nd trimester of pregnancy. If Target is able to find out which customer is pregnant before all the retailers, the marketing department can send them a tailored maternity offer early on, outlining valuable customer resources.

How can you accurately determine which customer is pregnant? Target thought that the company had a registration form for baby shower and started modeling and analyzing the consumer data of the customers in these registration forms, and soon discovered many very useful data patterns. For example, the model found that many pregnant women at the beginning of the 2nd pregnancy will buy many large packaged fragrance-free hand cream, in the first 20 weeks of pregnancy to buy a large number of supplements calcium, magnesium, zinc, such as good health products. Target selected 25 Typical consumer data to build a "pregnancy prediction Index", through which target can predict the customer's pregnancy in a very small margin of error, so target can send the pregnant woman's preferential advertisement to the customer early.

In order not to let the customer feel that the merchant violated their privacy, target put the preferential ads for maternity supplies in a host of other items that are not related to pregnancy.

According to this "Big data" model, Target has developed a new advertising marketing program, resulting in explosive growth in Target's pregnancy supplies sales. Target's "Big data" analysis technology is being extended from a pregnant customer segment to a variety of other customer segments, with target sales rising from $44 billion to $67 billion from 2002 to 2010, when Target uses "big data".

2.ZARA

Zara's average price for each piece of clothing is only lvhm One-fourth, but looking back at the financial reports of two companies, Zara's pre-tax gross profit margin is 23.6% higher than the LVHM group.

(1) Analysis of customer needs

At Zara's stores, cameras are installed in the counters and in all corners of the store, and the store manager carries a PDA with him. The aim is to record every opinion of its customers, such as the customer's preference for clothing patterns, the size of the buttons, and the small movements of the zipper style. The clerk will report to the store manager, the manager to the Zara internal Global information network, at least two times a day to send information to the Headquarters design staff, from the headquarters to make decisions immediately after the transfer to the production line, change the product style.

After closing, the sales staff checkout, inventory the daily goods up and down, and the customer purchase and return rate to make statistics. Combined with the counter cash data, trading system to make the day of the Transaction Analysis report, analysis of the day's product rankings, and then, data directly to Zara storage system.

Collect a large amount of customer opinion, in order to make production and sales decisions, such practices greatly reduce the inventory rate. At the same time, based on these telephone and computer data, Zara analyzes a similar "regional epidemic", in the color, version of production, to make the most close to the customer's needs of the market partition.

(2) combined with online store data

In the 2010, Zara set up online stores in six European countries, adding to the connectivity of vast amounts of data in the network. 2011, respectively in the United States, Japan launched a network platform, in addition to increasing revenue, online stores to strengthen the two-way search engine, data analysis functions. Not only to recycle ideas to the production side, so that policymakers pinpoint the target market, but also to provide consumers with more accurate fashion information, both sides can enjoy the "big data" benefits. Analysts estimate that online stores have raised at least 10% of the revenue for Zara.

In addition, online stores, in addition to trading behavior, is also the marketing touchstone before the activity product is listed. Zara typically conducts consumer opinion surveys on the web, and then retrieves customer feedback from the Web to improve the actual shipping products.

Zara sees the massive data on the network as a precursor to the physical storefront. Because the people who will search for fashion information on the Internet, they are more avantgarde than the general public in their preferences for clothing, the mastery of information, and the ability to generate trends. Furthermore, the consumers who will first learn about Zara's information on the Internet are also very high in the consumption of physical stores.

These customer information, in addition to the application in the production side, at the same time by the entire Zara (Inditex) Group of the various departments used: including customer service center, marketing Department, design team, production lines and access. According to these huge data, form the KPI of each department, complete the vertical integration spindle of Zara interior.

Zara implemented a large amount of data integration, and later by Zara under the Anglo-German group under the eight brand learning applications. Can foresee the future fashion circle, in addition to the table on the design ability, the table of information/Data war, will be more important invisible battlefield.

(3) Rapid data processing, correction, implementation

H&m has been trying to keep up with Zara's footsteps, actively use the "big data" to improve product flow, the results are not good, the gap between the two widening, this is why?

The main reason is that "big data" the most important function is to shorten the production time, so that the production side in accordance with customer opinion, can be quickly corrected in the first time. However, h&m internal management process, but can not support the "big data" the huge supply of information. H&m's supply chain, from the plate to the shipment, takes about three months, completely can not be compared with Zara two weeks time.

Because H&m is not like Zara, the latter is designed to produce nearly half of the rest in Spain, while h&m are scattered across Asia, Central and South America. The time of the transnational communication, lengthen the production time cost. In this way, even if the "big data" reflects the views of customers in different districts, the results of the separation of information and production will not be improved immediately, so the effectiveness of the "big data" system within the H&M is limited.

The key to the success of "big data" operations is the ability of information systems to work closely with decision-making processes, quickly respond to the needs of consumers, revise them, and immediately implement decisions.

(Responsible editor: The good of the Legacy)