According to this article gathered from "Make Data Your Killer App" keynote speech US travel data platform Lime thick head of Lu Chang (Alex Lu), published in 2019 Shence data-driven conference. This article will focus on the success of your Big Data nine ideas, contains the following:
Data Culture
Quality First
无快 Fuwa
Data civilians
- Decision-making loop
- Products and closed-loop operation
Data Security
Make good use of artificial intelligence
Leveraging the power
"The data make killer APP"
Smart phone era, everyone's phone has a lot of APP, we are using a variety of APP every day. APP Our lives will not become inconvenient. The era of big data, data on individuals like APP as corporate practitioners, we have to use every day. There is no data, work practitioners will become difficult. We talk to the owner to use the data for reporting the results, we use the data to help make decisions every day, our data product operations become more intelligent. We want to do the same data as did the killer APP, it is hoped the data into daily practitioners just need to APP.
"My first time"
I studied at Tsinghua is automated, a student of electrical engineering abroad. I first lasting bonds with the big data in 1997. At that time, AMGEN, this is the current market value of one hundred billion US pharmaceutical companies, not their own data team. We were invited to help them is to build a data warehouse. Project started at the other side attaches great importance. All project participants are engaged in their COO of speech blood boil. I was very poor English, but there are two words, or understand, the first sentence is "the data into a competitive advantage AMGEN" and the second sentence is "you can not do it boldly bad money."
By six months of efforts, a team consisting of a dozen people to complete the project. Follow-up project with stunning effect. AMGEN data warehouse to make smarter decision making, product sales increase very fast, but also made a lot of money. As a well-known pharmaceutical companies, AMGEN spend a lot of time and energy in addition to research and development of new drugs outside its core business is to sell drugs. They have a strong marketing team, spend a lot of money to advertise, according to the current word is "playing the brand." At the same time they have to spend effort to open up the way to the sale of pharmacies, insurance companies, according to the words now is "build channel." More importantly, they want to get the doctor network, because only a doctor can sale their prescription drugs, according to the current word is "to do the conversion." We build a data warehouse really become a competitive advantage this pharmaceutical companies.
In fact, before participating in this project I have done other projects, I did develop WEB page development, e-business, CRM and other systems. But I did not know where their own interests, pharmaceutical companies until after the project finished, I know, data is what I do. Although the technical challenges are many, but really attracted me was different data can bring, as well as real money value.
"Do Internet Data"
Later, at the beginning of the Internet we have joined Yahoo, becoming the first to do its data engineer. Although there was no large data argument, but it is indeed a substantial amount of data, and is constantly burst length. I summarize seven years of work at Yahoo in one sentence, it is "how to use the data to better understand the user to understand the thousands of users, hundreds of millions of users to understand and provide better service to them constantly," all for user.
Later, in an industry conference, I met later Google colleagues. At that time I asked him, traffic Yahoo is Google's ten times, but revenue was almost like Google, which in the end Why? He said, you want to know the answer, join it. So, I joined Google. But also because of this curiosity, so I understand Google's business product and profit model. I summarize the six years of work in Google in one sentence, is "to ensure that the user experience under the premise of how to use data and technology to make the search to bring more value to our customers."
After leaving Google, I joined Baidu. In Baidu know Wen-feng and Shence little friends, Baidu's six years, my work has always been around big data, cloud computing, artificial intelligence carried out, broaden their horizons. After leaving Baidu I joined Pinterest. Pinterest is a picture of a waterfall APP originator. In Pinterest, start the first practice for data used APP product ideas, so that everyone's company data everyday tools. "Do O2O + Internet data"
A year ago, I was out a cross-border small step out of the wall joined the Lime. Lime attempt to redefine travel, green electric scooters, travel to solve the "last kilometer" problem, where the data is everyone just needed a day. We do continue to follow the APP product ideas for data, and the data, cloud computing and AI overall consideration, in order to accelerate product innovation and sophisticated closed-loop closed-loop operation.
20 years ago, data value is art that requires opportune, so almost 80% of projects fail. Today, to realize the value of data to rely on technology, according to a survey in 2018, more than 80% of companies regard the data as part of its corporate strategy and AI focus, the moment has no one to ignore the value of big data. The famous investor Mary Meaker Internet Trends report in 2019 mentioned the success of the enterprise by 1995 is by artificial catch data for competitive advantage, 1995 to 2000, by virtue of electronic data as the Internet caught, gain competitive advantage , while since 2000 twenty years, is to use big data / AI technology to achieve more data value and gain competitive advantage.
At present, data / AI technology dazzling. So many options can also make big data practitioners screenwriters. Maybe the next big data and AI technology will be integrated into the head of several large platform. As Big Data technology platform practitioners, become part of the head platform has become the dream of many. But the technology has not developed there yet, still do success of big data challenge again and again. Next, I will share some thoughts about doing successful big data with you.
1. Data Culture
People always say that data is the CEO cultural thing, yes, without the support of the CEO, but where's the funding. But companies can not expect the CEO to understand big data. Head of big data, the company's various business leaders have a responsibility to the company's data values and vision clearly describe, to the CEO to see, but also to look at the whole company, to promote the culture of the company's data.
Read "The Lean Startup" book of friends all know, is the key to the data link ideas and practices and closed-loop iteration. Today, the scientific data is fire, using statistical and artificial intelligence algorithms from data found in the law can be applied to business and products, among which, the data is also key. Lime strong emphasis on cultural data. Product innovation loop, closed loop operation fine, in many O2O companies are very important concept. For Lime, the current product is green travel scooters, is to operate on a scooter at the right time to put in front of people. Because the data has become crucial two closed loop, using the data it slowly became a daily habit. Did not become a habit is very difficult to call it culture, culture is unconsciously you will practice something. Lime reason why so much emphasis on data, because we know that for start-up companies, no one can predict the future development of specific forms. But we believe - fast-learning to win. It quickly learn to rely on data, let us know where success, where failure to find out why, and constantly try new error correction. There is such a culture, the company's rapid growth also can be tracked.
2. Quality First
We are recruiting personnel data, they usually can identify a simple question whether the candidate really knows how to do data, this question is "how to solve the data quality problems." We should have heard of garbage in / garbage out, wrong data is worse than no data, there are many such examples, companies rely on a large number of decision-making data, but the data before the final decision found that the use is wrong, those caused by wrong decisions the negative impact of making decisions and perhaps even worse than no data. High-quality data representing the people's confidence in the data, only trusted, people can rest assured that the use of daily work, to become the new norm.
3. 无快 Fuwa
No fast break, we talked about a lot, Wen-feng led Shence data to make a good product in this regard. So that you can quickly build a data system, timely data and create value to you to use at the right time. Recently we talk a lot to do "data lake", its original intention is done quickly. Typically, after the release of new products, the intermediate can be used to retrieve data from the process through layers to the data. Lake destroyed the part of these data, then a user data query tool, you can use the data in a timely manner. Thus, whenever a new product launch, you'll be the first time point, using data validation product iterations effect. Its value late data becomes smaller.
4. Data civilians
When the data become the company's culture, when the data is everyone's daily needs of the company, we have to do with the idea of doing APP data. Each individual company can make use of the data in practical work. Data is no longer a "manager" of special tools. Data decision-making, the more people use, the greater the value. Data civilians is not easy to do. First, the data for the civilian population, that APP must do extremely easy to use, the threshold to be pulled down. Second, each person's way of data query, the results may differ. To answer a clear distinction between standard and free answers, reducing the risk of data errors caused civilian use. Third, from the point of view regulatory data, the more data the civilian population, the higher the risk of leakage. Strengthen data security and use practices is a must.
The decision-making loop
No company has a perfect world data. There is no one company owns all the data. Many of our decisions are made in the case of incomplete data equipment. So many times, say the data supports more appropriate. Data should not replace the dominant decision makers.
6. The closed-loop product operations
Small run, fast iteration has become conventional products do work, its core product test automation. This approach can also be extended to the operator. Product, for example, have an idea after making a rapid prototype, and then do the experiments on the A / B experiment, automatically generate statistical comparison of results and the next step is recommended. Here a lot of scientific data measuring methodology are automated. This is a large concrete data to accelerate product innovation and iteration killer APP, is already a lot of engineers, product manager everyday standard tools. My former club, before the experiment platform automation, one day only doing about ten experiments, automation can be done after the 1000-1500 experiment every day. Experimental platform makes the product significantly speed up the iterative loop.
7. Data Security
I do the first data item when it is covered to data security. Internet age, the importance of data security / privacy of the user and on a new level. Recently, the EU GDPR bill, the country also introduced a lot of new regulations regarding data security regulations. Compliance data era has come, do data must respect the privacy of users of God. GDPR accordance with EU requirements, if the enterprise leak user privacy, may 4% of total company revenue of fines. In January 2019, Google fined 50 million euros, Facebook is FTC fined $ 5.0 billion is an example. "Data misconduct" has been never an excuse for exemption.
8. make the best use of artificial intelligence
Not to mention the artificial intelligence of big data is incomplete. Often it said that the era of artificial intelligence, big data is king because it had not cut the relations between artificial intelligence with big data. A specific practice is to use artificial intelligence to improve data quality. How to maximize the value generated by the fusion of the two, we need to think about. Digital transformation of traditional enterprises in recent years most talked about topic. Overall, the business transformation of the traditional way is to collect more data of the road, the road is using artificial intelligence.
9. leveraging the power
In the end you should choose which platform? Privatization or public deployment deployment? Difficult to recruit people, not to build a data team? To answer these questions, leveraging the power may be a good strategy. Perhaps more important, especially for SMEs, start-ups, to seize the market than independent research. Numerous examples prove, standing on the shoulders of giants can make a successful big data. When using leveraging the power of strategy, but also to deal with the following challenges, first, cross-vendor system docking and optimization, second, and simplifies integration platform, third, data analysis and science, is still a powerful few can borrow, build team is inevitable.
in conclusion
Do big success and value of output data to be directly linked, talk about the value of big data bullying. Although the data is not a panacea, but is not the data is almost totally unacceptable. Data is fulcrum, with good corporate data may leveraging huge commercial value. Twenty years ago realize artistic value data, data on the value of today's more technical. I hope people will make the data's killer application, to make themselves invincible position in the competitive neutrality.
City trailer is about to start