Shulou(Shulou.com)06/01 Report--

This article introduces big data practitioners need to recognize what is the boundary of data analysis, the content is very detailed, interested friends can refer to, hope to be helpful to you.

With the advent of the era of big data, companies are accumulating more and more data, and the cost of obtaining data is decreasing. Many people begin to go to the other dangerous extreme, that is, everything depends on data, and any decision depends on data. As an operational data analyst, don't get caught up in the "data-only theory".

The database doesn't record everything.

Can you fully understand your business through the records in the database? The answer is no. To put it bluntly, the database only records the behavior that takes place on the business chain, but the result of the behavior does not represent the whole business.

For example, can you know the user's experience through the user's behavior data? No, we just guess the quality of the user's experience of using the product based on the "behavior results" of the user.

Real users feel that in their minds, it is difficult to reflect through the established use path and product features.

So, how to get the information that cannot be recorded in the database? The answer is actually very simple, through external means, to create conditions to obtain. It is summarized as "investigation" and "experiment".

For example, for questions that cannot be quantified, why not just ask them directly? The survey is divided into two ways: interview and questionnaire, each of which needs to be quantifiable.

The questionnaire survey suggests that it should be carried out regularly for a long time, and the data collected continuously are comparable in time dimension, and the value is far greater than that of a single questionnaire survey.

The way of experiment is also a means of creating data. Through the experimental group and the control group, create a comparative condition, and then quantify the difference, and finally form a reliable judgment.

Do not separate the influence of multiple factors

The most common mistake we make in our operations is to try to use a change in a "macro indicator" to assess the impact of an operational action (a change in strategy, a change in product, or a change in activity). Luckily, an operational action has a great impact on the business, so it can be reflected in the indicators. But most of the time, whether it is the change of strategy or the improvement of products, the impact on the overall business is limited, and the change of indicators is not sensitive.

On the other hand, the change in business indicators is often the result of a variety of operational actions, which can not be simply decomposed into "Aging Beverage C +". Or "A × B × C ×..." .

Some factors may magnify and influence each other, while others may restrain each other. Macro indicators are only the result of many influences, and the internal influence mechanism is a black box.

So what should I do? How to measure the impact of a specific operational action? The answer is the experiment, and only the experiment.

Data cannot represent logical reasoning

A person with confused logic, no matter how much data he gives, he will not come to the right conclusion. Whether we can form a correct judgment and reasonable decision, to a large extent, does not depend on the amount of data, as long as the data is sufficient (sufficient information).

Most of the time, what we really need to exercise is our own ability to analyze problems, or logical reasoning.

Whether you are a wise decision maker or not depends not on how much data you have, but on whether you can interpret the information correctly from the data. We need to remind you that when the data reaches a certain amount, the more data, the greater the possibility of making mistakes.

Most of them become lazy because of the data.

Why would you say that? Most managers will give up thinking more or less because of the abundance of data resources.

Many people, because of their long-term business experience, actually have good intuition. Now, however, he is led by the nose by a data analyst who is not familiar with the business and can only do some statistics from the database. Many people are "in awe of data" too much. This phenomenon is worthy of vigilance.

A senior entrepreneur once said that data analysis makes people short-sighted and even blind.

Therefore, really good data analysis (business analysis) has a very high threshold. This threshold does not come from the application of analytical methods, but from the understanding of the business. Only when we have a deep understanding of the business, can we use the analysis method to the right place, interpret the information correctly and get the conclusion.

Big data practitioners need to recognize the boundaries of data analysis to share here, I hope the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.