When can companies really profit from big data?

Large data contains great value, I believe that everyone who is concerned about it technology innovation will hear this sentence. Yes, from the http://www.aliyun.com/zixun/aggregation/13584.html "> a large number of data to filter out useful information, and then through various means to transform information into insights, so as to make the right decision to promote business development. In such an information chain, the enterprise needs to ensure that every link is not wrong in order to transform the data into value. But how many companies can really do that? Few! Big data is very hot, but when will the big data really bring profits to the business? In this respect, data analysis and marketing expert Brooks Bell published her opinion on Techonomy.

The McKinsey agency says big data will be "the forefront of competitiveness, productivity and innovation for the next generation of companies". But the reality is that many businesses and managers are beginning to blindly collect data and analyze it, expecting a quick return. Unfortunately, they failed to do so. Most businesses are less than 108,000 miles away from the data, which is more than just a lack of appropriate technology. To make big data truly impact corporate profitability, there are three deep-rooted challenges to address.

First, the way to make decisions is still common. In the business world, the opinion of the "supreme right" has a great impact on decision making, and this is a very common phenomenon. This is a common problem in many enterprises, large data can be corrected. However, the real need to change the concept of enterprise, leaders in making decisions to get rid of the "head" of the bad habit, so that the real data to speak. Just collecting more data will do little to overturn this mentality and even make the process of changing the mindset harder.

In a recent hot-Signal bestseller, "Signal and Noise", author Nate Silver said, "If forecasters don't trust people, then people won't listen to the weather, even when they really need them." It's like the "wolf" problem that exists between the CEO and the data. If the analysis is wrong, or worse – the data has not been collected correctly from the start, then policymakers will surely lose faith in the information and the employees who provide it, and return to the "head" era again.

The second challenge is the lack of talent skills. For the moment, people who can play big numbers are far from satisfying the needs of their businesses. Wanniwal Bush, the father of Silicon Valley, Vannevar 70 years ago: "There will be information-seekers in the future who will look for clues in a lot of ordinary records and eat their own music." However, according to the McKinsey report, there are currently only 190,000 highly trained data analysts in the United States that are far from meeting the needs of the big data age.

According to a survey by the SAS Institute and the IDG Agency, 57% of participants said they lacked qualified skills and experience in data analysis. Lack of confidence in analytical tasks is only part of the challenge, and employees working on data need to do more to collect the right metrics with enough precision.

Business managers need not have to recruit a group of data-scientist-level elites to report directly to them, and they will have to encourage them at all levels to nurture analysts, impart core skills, best practices, and be as precise as possible in the process. This would increase transparency, encourage demand for data and help disseminate essential skills.

Knowing how to work with data is the third challenge. Even after addressing both of these issues, it's a matter of figuring out what kind of business can gain from large data. If it is not possible to guide action, it is meaningless to collect more data. In fact, gaining insight is on the one hand, and practicality is also a sign of analysis. Can companies get predictable predictions and forward-looking decisions from the "noise" of a lot of historical data?

For example, a handset manufacturer may be able to collect a large amount of consumer data, unless it can be applied in practice to improve the customer experience, otherwise it will only have theoretical value. For example, a chain of retail companies through accurate mail marketing to obtain customer information, but if the sales department does not use this information reasonably, then the opportunity will be fleeting. The culture of data must be disseminated to every employee of the enterprise if large data are to be successful.

Not only in the big data age, the "discomfort" of the data is the main cause of this problem, the big data just magnified the problem. The novelist, Borges, has described this issue in the bar library. The universe is filled with countless library shelves with books that look the same, and each book consists of letters and punctuation with different random combinations. In this library, all ideas and events are recorded, but any insights are hidden in countless nonsense. The librarian in Borges ' works can't make use of this huge resource and can only muddle through. What is the difference between the application of large data and the enterprise?

Nate Silver also mentions this, and he thinks that distinguishing useful signals from noise requires both scientific knowledge and self-awareness: calmly accepting things we cannot predict, and the courage to predict what we can predict, and the wisdom lies in how to discern both. In other words, the data cannot be used to uncover the truth, it can only make assumptions, and then we prove it through repeated tests and practices.

Big Data gives hope. But what we need to do is understand the importance of data and then make the most of it at every stage of planning and at every level of the enterprise. It is necessary to master the sufficient conditions for small data deployment to make good use of large data. The focus of business should be to allow more employees, more regularly, to make better use of the manageable data. Then let the business gradually be based on data to take action, only in order to make big data dream become a reality.

(Responsible editor: The good of the Legacy)