Big data, don't drag the sky!

Big Data This thing, http://www.aliyun.com/zixun/aggregation/11892.html "> High-end atmospheric grade. If you're a businessman, when chatting, don't read between the lines of your strategic plan that covers the plan for the big plans. Ten, will be peer contempt. Feelings you are a winter is coming, but do not know the reserve grain of silly a.

Now, more and more media, whether it is government media, commercial media, or the self media. Anyway, with a mordant. Chatting up large data, all with admiration and awe. Almost bowed to the big data. That argument, just like a few years ago, we hype the concept of electricity quotient. Almost resolute and solemn shout, the future does not electric trader-die!

As a result, death is a dead bunch of people, but at the same time, the concept of "electric quotient" died only one breath. Why, because we found that the electrical business is not completely independent, he is inseparable from the real economy. Therefore, in the general "electric Quotient" concept, we have a more refined version of the saying: O2O, think this is the electrical quotient. The pure thought that the electricity business is to put the shop online, is not culture.

But is this the end of the world? Don't forget, after O2O, we are brewing a new gadget, that is the fan economy. Now it has begun to shout out. The future of commercial competition, no fans die, fans live, the situation is much like the electricity business slogan. Whether you are an online shop or a real store, whether you are a giant or a new entrant. Fans, fans, everything is a fan. With fans, all the way, no fans, you are tomorrow's Nokia, but also meet the Apple after the Nokia. I don't know what that means. It means, you are a giant, no fans, you also have a stunted big head doll, in addition to die, is to be acquired!

So what does this have to do with big data?

To tell you the truth, it's not much of a relationship, but don't rush to shoot bricks. Mechanism is similar. Today, we talk about big data, just like a few years ago, we talked about "electricity quotient". is an extremely general concept.

To a certain extent, no matter the business environment or political environment, the brewing of large data, even not to let everyone from the pan-concept of electricity to realize the "O2O".

Why? Because once the "General Electric quotient" concept, his good or bad, is directly reflected in the business behavior. So, after the shock. Immediately found that pure electrical business is not feasible. Must be O2O, online under the need to cooperate to form an ecological complementarity, and really inspire their own potential.

And the big data, he is more ambiguous than "electric quotient" this concept, more not qualitative.

For instance.

Now a large electric dealer's electronic statistics clearly show that the country's large regions of the strong demand for salt, and in accordance with the growth rate of demand, the electricity dealers must immediately replenishment. Otherwise, salt will be out of stock.

As a decision maker, are you filling or not?

You know, according to normal understanding, salt this commodity is no reason for large-scale sales soared, but, the system summarizes the national user's purchase information to draw the sales figures are basically not wrong.

But the problem is, if according to the future trend of sales figures to purchase salt, but sales suddenly decline, then, the cost of logistics warehousing is not small.

What does the case reveal? Large data in the use of the process, has his natural short plate. What is this short board? He is not big enough. As the case says, there is no reason for mass sales to skyrocket in the normal condition, but the charts provided by your sales system clearly show that this time and future trends in salt sales are growing. And you, even if you see the data, also flatly do not dare to make procurement decisions.

Why, because even if the sales figures give you a possible growth message, it goes against the logic of the pre-judgment in your mind. So you will hesitate. Unless, the big data can be "big" enough to cover all the variables.

So let's move on to a different case, from another perspective.

According to a company survey, China's average per capita wage rose by 7% per cent a year, barely coping with inflation. This data is accurate and reliable, is the company through the whole country payroll income summary, by the system actuarial

I wonder what it feels like to see this message? If you are not a government official, just a wage earner. I'm sure you'll feel a lot of nonsense. Why. Because what you care about is not the percentage of the wage increase. But how to get your wages up. Because, if you want to calculate how much salary rises, you own these years to receive the money subtraction, lightly loose even if come out.

And this reveals nothing. Reveal the relative "invalidity" of large data. Even though you can draw something with large data, what you come up with is worthless to me. So, the big data or the small data. One more glance is a waste of time.

(Official agency statistics "dining out" crowd more and more information, the partial location is not so good small restaurants useless is similar to the mechanism)

So, if the above two cases are summarized, we objectively summarize the defects of large data

One: If the associated information of a transaction can not be fully accommodated into the "large data" system, the reliability of large data, the usefulness will naturally weaken.

Two: A lot of things, we don't need big data to tell us. Because we knew. What is troubling us is how to solve the present predicament. And the big numbers are obviously not helping us in the short to medium term.

So, to a certain extent, we can say: Big data, that is, a relatively high reference value of data. If you are in a business with too many variables or if you already know the cause is just a lack of capacity. Big data is not really that valuable to you.

People who are always on the big numbers, in fact, probably don't understand the workings of big data.

But at the same time, in order to make everyone's perspective more comprehensive, give a positive case.

Which region of China is the best woman in stature?

The answer is: Inner Mongolia

Why, according to the data of a large electric dealer platform. Women in Inner Mongolia buy lingerie sizes that are more "plump" than any other area in the country.

In business applications, this data can help the planning of the establishment of women in Inner Mongolia underwear store businessman, ahead of the market, the shop opened to Inner Mongolia, only to find their sourcing in the local due to the size of the wrong and completely sales do not go out

Of course, why should I put this case in the end?

Because, although this case is a positive big data application case, but can reverse to large data not so good interpretation. What's the solution?

If we take a closer look at this positive case, we will find that this case can be applied on the basis that women's stature will almost never change significantly after adulthood. Moreover, this case is only relatively focused on the merchant's purchase of goods on a single, and does not involve other factors such as competitor information, local operating costs, etc.

What does that mean? means that even though large data can be applied, it is limited to the relative "variable" of the local level, otherwise, his energy efficiency is far from matching his "big data" appellation.

Therefore, to some extent, we can say that large data will show more and more value even in the future, but because it is too extensive, most business people or individuals do not need to be so nervous about it, except in some industries. Unless, you are using large data to analyze some kind of "relatively narrow quantitative". Otherwise, large data in more than 10 20 years. In the relative subdivision of the industry, can not say chicken ribs, but want to play a big value, it is difficult.