All kinds of marketing methods have long been dazzling, but their essence is to study customers (consumers), study what customers want and need, and make products or services targeted. The era of big data has given it a new term: precision marketing. The first fields of application of big data are mostly customer-facing industries, and the first scenarios are mostly precision marketing.

"wine is good but afraid of deep alleys", the information of products or services must be delivered to customers to facilitate transactions. It is generally believed that to convey product or service information to customers depends on advertising. Advertising has been around for a long time, and the guise of "three bowls of wine are not enough" is advertising. In the era without the Internet, we are familiar with TV advertisements, radio advertisements, print advertisements, out-of-home advertising signs, etc., and of course, yelling and selling. But in the past, advertising was one-sided and did not distinguish between audiences. Later, merchants collected customer information and had CRM. After customer classification, they could better serve different customer groups. The Internet + big data era has opened up new development opportunities for CRM. Managing customers is no longer a simple matter of digital statistics and direct mail and fixed investment without personality (or simple clustering). As merchants know more about customers and have a deeper understanding, they have the opportunity to provide customers with personalized marketing solutions, further improve customer experience, and become personalized marketing or precision marketing. The era of big data has made many impossible things possible in the past, and marketing activities have also won new development opportunities.

The era is different, and the form of business operation will change, but the essence is two things: open source and throttling. Open source is to open up new customers and discover new business opportunities; throttling is to reduce internal operating costs and improve resource utilization efficiency. To achieve all this requires data-based decisions. In the past, people also collected and used a lot of strongly relevant data related to business activities in long-term business activities, and also formed the criteria for selecting customers. In view of the technical bottlenecks at that time, the cost of data collection and data analytics for large samples was too high to be widely used. In the era of big data, it is possible to collect and store data cheaply, and cheap computing resources make data analytics possible.

Behind the precision marketing of big data, it is used to observe customers with multi-dimensional data, describe customers, that is to say, portray customers. It is not an exaggeration to say that "relying on big data allows marketers to understand customers better than in the past, and understand their needs better than customers themselves". Marketers do not want to know who customers are, where they are, what their consumption habits are, what they need, when they need it, and how to communicate information to them more effectively. Through data collection and data analytics analysis, the answer can be found. Precision marketing can not only help merchants open source --- find potential customers, but also help merchants cut costs --- find latent risks. When we know more about customers, we will know which customers may be at risk in our business.

If you ask each operator whether they will use their experience to market, most of the answers are yes. But if you ask the operator whether they will use data for marketing, I am afraid the answer is varied. It is generally believed that the application of data for marketing is a matter for large companies, and it is not relevant to small companies. In fact, large multinational companies, small street vendors, use data for marketing, will receive unexpected results. Don't believe it? Street vendors pay attention to the weather forecast (windy, rainy, or sun exposure) to know what business opportunities are available tomorrow, and then know how to stock up. It is recommended that people in small and medium-sized companies do not reject the concept of precision marketing, but may wish to learn the thinking method of precision marketing. Even if the operator has rich experience, it will be very helpful to digitize the experience to the operation.

The book "Subversion Marketing" is to teach readers how to use big data to do marketing. The book is rich in cases and highly readable in language. It is worth reading for friends from all walks of life who are concerned about big data marketing.

I agree with many points in the book: "Big data redefines the rules of industrial competition, not the size of data, not statistical technology, nor strong computing power, but the ability to interpret core data". In today's world where many people are torn by the definition of big data, we should indeed pay more attention to the core value understanding and application of data. The "asking the right questions" raised in the book is also very important. Operators must have a lot of questions at ordinary times, but when they are asked, there may be deviations, resulting in "a thousand miles lost". Improving the ability to ask the right questions involves thinking and methods, which need to be improved through exercise. Verifying whether the question is asked correctly is precisely where data analysts can contribute.

The book also raises two questions worthy of deeper consideration:

It is not enough to simply discover the consumption habits of different customer groups and remind customers to spend in a timely manner. For example, the normal rational consumption of a consumer is at the level of 2,000 yuan a month, which is generally consumed in two stores, A and B. Store A uses the concept of precision marketing to make consumers spend all the 2,000 yuan in Store A, and as Store B catches up, consumers may return to Store B to spend the 2,000 yuan. In today's oversupply and insufficient demand, the existing consumption amount cannot be distributed or migrated among different merchants to increase the total social consumption. A higher-level application of big data marketing is to know in advance the needs of customers that have not been met or even discovered. The value mining of big data has the opportunity to connect merchants (including manufacturers) and customers, allowing merchants to provide more products or services that meet customers' personalized needs, and increase customers' willingness to spend. This is a new challenge for data value mining workers.

Is more data really better? Many big data companies are keen to use crawler software to "crawl" all kinds of data on the Internet. However, the value density of the same data set is different in different application scenarios, and for specific application scenarios, it is not that the more data dimensions are the better. It is necessary to collect data and use data around the application goal. Increasing the dimension to collect more data will definitely help to describe things in more detail, but it will undoubtedly increase the complexity of processing data. Every technological advancement brings new imagination space to human beings. It is inevitable that the desire will expand and the confidence will be full, and the cognition of the world will also increase in dimension, even uncontrolled dimension. After that, it is found that the dimension increase brings the occupation of resources, and the wisdom cannot keep up. Uncontrolled dimension increase will complicate the solution. If you calm down, you will restart the dimensionality reduction thinking. Perhaps human cognition and wisdom are in ascending dimension, reducing dimension, and then ascending dimension, and then descending dimension alternately forward. The dimensionality reduction thinking in this book, when necessary, returns to the original thinking to inspire people.

Tools and means are important in the era of big data, but thinking methods are more important.