[Post] Bi inspiration: doubt creates value

Author: Region name from: it. icxo. com

It is no longer appropriate to use Google's single answer option to compare the charm of a business intelligence system. If the query returns one and only the correct answer, you will feel very lucky, because to be honest, there is very few correct answers to any bi questions.

Excellent Bi is just the right direction for you. Imagine: 52% of consumers who choose gray trousers also choose white shirts, but if blue sweaters are also available, only 31% choose white shirts. Therefore, if you are a seller and want to sell a large number of shirts, you will find a way to ensure that the blue sweater will not appear in the eyes of consumers. Even so, you only have more than half the chance to sell a white shirt.

If the seller knows that the shirt is blue, the proportion of consumers can be increased from 52% to 67%. What will it do? In this case, he will produce a lot of blue shirts. But what if he knew that the proportion of his shirt would suddenly drop to 14% if he sold it to consumers? He is hard at this point, especially if he cannot control the number of sweaters.

If he is desperate, he will bet on his blue shirt and want a limited number of sweaters. But if he is conservative, he needs more information before making a decision. What are the overlapping parts of shirt and sweater sales? Does the sweater color have a relationship? Does the daily time period have a relationship? What is the region? Conservative Bi users will constantly query the system until the best solution appears.

This approach seems reasonable, because the more you know, the more wise you can make decisions. Is that true? Not necessarily. To be exact, I agree with the German philosopher Goethe's statement: "The wider the knowledge, the more doubts it has ."

However, most end users have the opposite idea. They expect a simple answer. It would be better to get the only and correct answer like Google's "good luck. If not, end users may be annoyed, confused, or even skeptical about the benefits of Bi.

When Bi is used only by experts, this is not a big problem because experts know how to use Bi software to calculate accuracy. But as described in the article "Bi is more common", Bi tools are now adopted by more and more employees, but they are not statistical experts.

The masses do not know the importance of suspicion. However, Bi is actually serving commercial applications through suspicion. Bi forces us to draw skeptical conclusions, return to the preliminary stage, and reset assumptions and conditions. It promotes us to think, not to give facts.

Bi allows you to query a series of related data from different perspectives, so that you can get closer to facts under a set of conditions, which is far from providing facts. A large number of end users must understand this before starting to use bi. Otherwise, many wrong decisions may be made, causing heavy losses to the company.

Therefore, end users must receive Bi training. It is not only about how to train employees to use tools, but also about why to use bi: You need Bi tools to help make the best decisions, rather than making the right decisions.

We often analyze too much information in depth to obtain the correct answer to a question. However, we often encounter a situation where the more information we obtain, the less luck we get.

In a book titled brain organizations, Moz rubinstan and Alice fileltenberger wrote that in the 17th century, a European gentleman "was not as informative as it was today." I am afraid that I can explain why the gentlemen at that time were so keen on their own world view that they led to some meaningless moves: Launching War and sprinkling white powder on wigs. They don't doubt their ideas because they don't know enough.

In my opinion, this is the essence of Bi tools: it is not to convince you of what you know, but to encourage you to think more questions and questions.