Big data: The digital dream of the boss of Procter and Gamble

Meng Problem:

Q: I know what data is, but what is big data?

A: Have you Seen "Sherlock Holmes"? Sherlock Holmes was always able to draw accurate conclusions by observation, and in others ' eyes he was as capable of prophecy as God. In fact, he was merely combining and analyzing the observed information and drawing conclusions through reasonable deductive reasoning. For example, he will analyze the victim's nails that greasy hair silk is the life habits of people, and the soles of the mud in the cracks in the surrounding areas will have. In the age of large data, everyone can become Sherlock Holmes. You can imagine that business behavior is defined and recorded as data, and then people make predictions, judgments, and actions based on data analysis. The big data presents a lot of things that we can't see now, to help us decide that future human and corporate behavior will be more rational and efficient with large data.

Q: Oh ... Where did the data come from?

A: The data comes from anywhere. The big data age, the whole world can be seen as being "monitored" at any time, you read a book, eat a meal, can produce data and be recorded.

Q: How can this be! How can my personal data be easily recorded by others?

A: Very simple. Even now, some of your data has been recorded. For example, you buy clothes on Taobao, to choose their own size and favorite color, the data in your order after the submission to the merchant, after accumulating a certain amount of data, the merchant can predict the next time you may buy what kind of clothes. If you don't choose your own, you will be pushed forward by the satisfied product.

Q: Is it really that smart?

A: The richer the data type, the greater the amount of data, the more reliable the results of analysis. In addition, the data is ubiquitous everywhere, but in the big data age, people will consciously use more convenient technology to collect data and use it.

Q: Is that such a big data age, not so much like the so-called cloud technology? is to upload the data and share it.

A: Well, the cloud should be counted as a support system for large data that allows data to be transmitted and shared, but the essence of large data is analysis and application.

Q: The big data is limited to the Internet, if in the country, where is the data?

A: As I said before, the data is everywhere. For example, in rural areas, temperature, humidity, soil composition These are not data?

Q: Are these data collected by farmers themselves? How can these data be widely disseminated and shared after collection?

A: So large data will generate a lot of business opportunities, in data collection, collation, transmission, storage, calculation, analysis and application will appear more opportunities for money.

Q: Do you need money to share this data?

A: This depends on the specific business model, but the data is definitely the most valuable commodity in the future. Yesterday is not still a website text to invite you to Blind Date? You see, your phone number has been sold, this is the primary stage of the big data business model.

As CEO, I have a mission: to make Procter and Gamble the world's most technologically-content company. To this end, we have carefully studied the broad application prospects of digital technology and analytical science at all levels of the company's business--from the company's R&d labs to new materials, to maintaining good business relationships with retailers, to producing products, building brands and communicating with consumers. The benefits are obvious: innovation has been effective, production has increased, costs have fallen, and there are expected to be faster growth in many ways. Build enterprise Data "Lighthouse"

When I was a manager in 1984, I got a tape of consumer feedback from the customer service hotline and listened on my way home. After returning to the office, I will read and reply to the feedback received. Today, these methods are clearly not enough-because there are blogs, microblogs, and various mediums.

So we created a "consumer pulse" thing, using the Bayes principle to analyze the massive feedback data, classify them according to different brands, and send them to the relevant principals. I am personally responsible for the feedback of the P & G brand. This helps us react to the market's movements in real time, because we know that if someone says something on a blog and you don't respond in a timely manner--or, worse, you don't know--then things are out of control when you find it wrong. This technology also allows us to improve the work in progress. For example, we are introducing a new laundry aroma, called "Downy unstopables". Real-time feedback from consumers on this product can help us discuss how to promote the market in the most effective way.

On the basis of practicality, we also believe that only the continuous improvement of products to be successful, the use of digital technology, which is also easy to achieve. So we set about digitizing the various aspects of the company's business-from manufacturing to shopping malls. We believe that digitalization is an important source of competitive advantage.

For example, in our production system, we create a system that allows people to use the ipad to download data from the production line in real time and share it where we upload the data. We have not yet reached this point, but we can foresee that in the future, anytime any product in the production line of real-time data will be able to appear in my notebook. At the same time, I want to see the cost of the product at the same time. This is a bit difficult at the moment because the current billing system is not designed for real-time production and is typically used to analyze past data. But we are working to integrate our business systems and financial systems to meet our requirements.

In transportation and logistics, we have designed a program called Control Tower. It allows us to see the current transport situation: input and output, raw materials and final products. Our use of trucks is likely to be second or third in the US, and through this technology we have reduced the waste of about 15% of our transportation resources and reduced our cost and carbon footprint.

Not only that, we also want to build web-based links with retailers. For example, we support the GSDN (Global Data Synchronization Network), which is an important standardized database that allows us to engage with our retailer partners fully automated business without the need for labor. The GS1 Industry Association, which did a study a few years ago, found that 70% of the orders of retailers and producers were faulty. But if everyone starts using a standardized database like GSDN-data will be saved exactly-then that number will fall to zero and save tens of millions of dollars in collaborative costs.

The other thing we're doing is using our power to bring cutting-edge technology to retailers that can't afford it, such as the Philippines. I used to do a little shop there, and we were able to offer high tech order systems to help people run their business better than they used to. Retailers can use some of our mobile apps to place orders on our wireless networks, or if they don't have a wireless network, they can bring their phones back to the office for an order. Everything is quite convenient.

We have also set up enforcement standards to enable mobile phones in some developing country retailers to form images. We believe that any retailer will need to use some means to increase sales as much as possible. For example, if you have a store and you support this standard, then you can open the program on your phone, lift your phone, look around the store, and compare the program to what you see in the store. Finally, I want the system to take a picture of a shelf, use a computer to compare, and then automatically send the advice back to the retailer to help them rearrange their goods so that they can maximize their sales. We are working on this.

Data models, simulations, and other digital tools are reshaping the way we change. Past research means a lot of time and effort to build consumer groups-you need the right race and age distribution, and other factors to make it representative. Now, the collection of available data is so large that a direct definition can be made of a representative group.

For example, when you design a traditional method of urine is not wet, often the prototype design just completed, it has cost tens of thousands of costs, not only that, the design is entirely by hand. Now, with modeling and simulation, you can repeat the design tens of thousands of times in a few seconds as long as you get the data, so this is the advantage of Procter and Gamble. We are engaged in production in more than 80 countries, our products are sold to almost all countries, we contact more than 4 billion customers daily. Imagine the amount of data we can design for any baby in any part of the world. Data, brand "adrenaline"

Every Monday morning we have meetings with management teams from around the world-sitting together or meeting remotely. We summarize the business of the previous week and look at the data. Everyone has to put forward their own opinions. This is in real time and is in the long run. This allows us to discover patterns, make decisions, and achieve them.

Because we meet every week, the data source is a challenge. I also use the Philippines for example, if we buy joint data from a company, and the company does a questionnaire in the Philippines every two months to get data, then every Monday will become irrelevant and the quality of the data will not be high. So we've been working with data companies to help them understand what real time data means to us. This is a hard rule for us--understand the problem of the data and push it all the way to the data source and replace it.

We think that data is worth less than a brand, but data can help create a brand and keep it alive, so the data source is very important. Therefore, we will do our utmost to protect all our consumer data, which is an enterprise-level risk management problem. We strictly separate the data of different retailers and set strict rules. For example, when working with different retailers, how long the "cooling period" is appropriate, which is consistent with our strategy to become the most digital company. We cannot do this without being the leader in the privacy of data security. We need people who understand computer modeling and simulation, and we need someone who is really proficient in computers, whether it's basic coding or advanced programming. You only have done simulation, will understand the importance of data, is a famous saying: "Crooked." ”

Although Procter and Gamble are doing well in analytical thinking, we will recruit good people and train them. I still remember the first day I went into the company and a manager said, "Throw away the MBA textbook, we'll teach you, we'll get you an MBA." "I think this sentence is still useful today." But analytical thinking has become extraordinarily important. So we need revolutionary ideas, and these revolutions must be driven by data.

Reviews One:

Switching platform triggers large data age

Wen/Xie, it commentator

Even though the arrival of the big data age was discussed half a year ago, it was a topic that was discussed in a professional circle. To this day, the concept of "big data" has been widely disseminated both inside and outside the industry and in the mass media, and the efforts and innovations in this direction have been emerging. From different angles to see large data, it may be a big opportunity, great development, great innovation, but also may be a big crisis, great destruction, big elimination.

At present, the most active areas are network terminal innovation and network infrastructure innovation, that is, the so-called large data industry chain of the foreground and backstage. From the well-known desktops, notebooks to smartphones and tablets, and then to the upcoming network television, network cameras, network glasses, as well as the discussion of the network of light bulbs, auto and a variety of strange network terminals and sensor systems, the material world and human society more comprehensive, Work is progressing smoothly and rapidly into the data world. From the familiar traditional cloud computing and data centers to today's public cloud, private cloud, open cloud, closed cloud, to an endless stream of hardware, software, data storage and analysis tools, the background of large data is being serviced from software level (SaaS), Platform-level services (PaaS) toward infrastructure-level services (IaaS).

On both roads, there seems to be no great theoretical or practical obstacles that can prevent this process.

The real battle is still in the big data of the middle, that is, the network platform, this big innovation is the Big data era really come to the tipping point (Tipping points). No matter how colorful the foreground is, no matter how strong the backstage, after all, there needs to be a system, a framework, a service to connect people, objects, and people to the data produced by natural logic and social logic, docking up, integration, can release the potential economic and social value. This connection, docking and integration of the way users like, the lower the cost, the higher the efficiency, the more data, the greater the value of the platform, in the large data ecological circle in the higher position.

On the current situation of industrial development and the internal needs of large data times, the next 35 years will be in the network platform level has the opportunity to produce innovative breakthroughs in the following three major directions:

Personal data integration. This is the natural deepening and expansion of the Web2.0 Revolution, the ultimate goal is to create the real "data Man", that is, the individual-centric, their behavior on the internet and everything in the world about this person's generated data together to accurately describe the protection of privacy under the premise of intelligent and personalized service matching. In this respect, Facebook and Apple have the best foundation and go furthest. A series of new concepts such as my data, self quantization (quantified self), and "nanotargeting" are emerging in the industry, and a group of services and mechanisms around the acquisition of individual complete dynamic data are being tried.

Public service data integration. The network public data service which has lagged far behind the development of the Times has sprung up in recent years, from the fragmented, backward, rough and passive state to a new stage of integration, dynamic, refinement and initiative. The Government Data Service website, represented by Data.gov, is becoming a new and powerful force in the era of large data, which expands and enriches the space and depth of Internet service, under the impetus of legislation, budget, supervision of public opinion and public urging. The level of development and competitive strength of a country, a society and even a city will be closely linked to the level of its public service data integration and service. The level of public service data integration will soon become one of the main signs of "soft power".

Material production data integration. The design and manufacture of material products have been far away from the internet, and are now converging with the network industry with great speed and great power. The third industrial revolution, represented by "3D printing", greatly improved the imagination of the network world and the data world, and greatly expanded the industry boundary of the network industry. In the past, the network industry can only carry out fully data products and services, or through the network platform to help the material products and services to promote sales. However, the emerging networked and data-producing models show that the transformation process from data to physical objects begins to enter into a new historical stage of low cost, large-scale, breaking time limit and individuation. This will redefine the industrial chain and business model of many product manufacturing industries, integrating the data needed for the design, manufacture and circulation of material products into the upstream industry.

These three directions are just a space composed of three dimensions of personal, social and material worlds, which are organically integrated in the era of large data and create opportunities for industrial development and social progress. Any significant progress in any one or three dimensions of this space will be the gospel of the large data service industry. This is not an imaginary myth, but a visible future. Reviews two:

The value depression in the era of electronic business data

Wen/Zhang Jianseng, former Jingdong mall data analyst

2010 years later, the "cloud data" concept broke the data time, space constraints, the Big Data era door is open. Several large domestic electric dealer sites have more than tens other active users, and offline than they have more easily access to consumer data, commodity data characteristics. Jingdong Mall daily average turnover of more than 100 million, order volume of more than 500,000, the enterprise has a complex operating process, these are data can play a major role in the link, the full use of data can be extremely high ground to improve efficiency, cost savings, its manifestations include the following several aspects.

Data to help you make decisions

The past operations drive data will become data-driven operations, large data not only refers to a huge amount of data, but also contains data segmentation, the enterprise almost all the links will be in the form of data to show, such as the time node of each business link derived from the efficiency optimization. On Amazon, there are a lot of operational reports and data processing every day, operation strategy, marketing strategy change is mainly to look at the data, its own definition of automatic replenishment model is based on the time series and the principle of extreme value formed, effectively solve the total reliance on artificial orders, replenishment mode, improve the efficiency of inventory management.

Analyze the user's intelligence

The most fundamental thing is to do the user experience, when the electric trader has a large number of consumer purchase behavior data, consumer research can even specific to a certain user, including regional purchasing power, commodity regionalization, customer layering, shopping cycle, shopping bias, the reasons for complaints and many other data, The combination of indicators will provide an important basis for enterprises to implement differentiated strategy and precision marketing. Through the data analysis, can also effectively identify with the competitor difference factor, creates the new blue sea, provides the more suitable shopping experience for the consumer.

Create a three-dimensional "data network"

The internal information flow of electronic commerce can be transformed into data, multidimensional, multi-angle use of data, through a core dimension of the scope of the data gradually expanded, the cause of a behavior and rationality through more than 10 or even more data standards to show, to make it more accurate and focused, For example, sales data can be sold as the core, the product sales of regional, cyclical, after-sale return, customer complaint rate, order periodicity, customer loyalty and other indicators of comprehensive analysis.

Let the data "see, Touch."

The traditional data analysis is more in the form of simple charts or PPT, not intuitive enough, since 2010, the emergence of data information maps, for data analysis and results output provides a very good visual effect and understanding, using a simple graphic combination of a single chart into richer content results, Greatly stimulated the people's sensory nerves, so that the boring data become vivid image, data information map is only a manifestation of the in-depth development of data visualization, large data age will be derived from a number of similar methods.

Big platform needs "keeled"

In terms of time, the basic analysis is based on historical data and real data, the model can provide long-term prediction data and evaluate the rationality of real data, they complement each other, and the complementarity and contrast between different methods can provide more accurate reference for business development. With the stability and maturity of the electric business model, the use of the model will gradually increase, especially in consumer research, sales forecasting, inventory management, simple or complex methods are necessary, the role of the two are different, in the construction of large data platform, the electrical business needs to better balance the relationship between the, so that it plays a corresponding effect.

Everyone for me, I for everyone

Electronic Business data is now difficult to obtain, some of the public data its accuracy is doubtful, the competitor's analysis is also based on the objective, limited the entire industry to the reasonable use of data. The establishment of a large data concept, improve the value of the enterprise data, the enterprise part of the function is also changing, the data generated services are emerging, such as Amoy nets, Taobao, such as the regular release of Internal price index, category Sales report, is the internal data sharing a good beginning, Many enterprises will be combined with their own and industry open data on a field of electrical business to carry out specialized research, for new entrants or the development of the industry to provide in-depth services; In the internet age, data sharing is an inevitable trend.