Jeff Bezos: The future of business is machine learning, a focus on customer interests and quick decision making

The founder and permanent head of Amazon, Jeff Bezos, is quite well versed in how to build a prosperous business from scratch. His company continues to grow, and its capitalization is increasing year by year. Analysts believe that Amazon will be the first company in the world with a trillionth capitalization. It is clear that Bezos himself is not in trouble either. He occupies the second line in the ranking of the richest people in the world.

In a recent letter to shareholders, which was sent a few days ago, Bezos explains why he considers it necessary to act one way or another. In his opinion, for example, focusing on customer needs is the best way to achieve success. He also talked about how Amazon uses machine learning and artificial intelligence (its weak form, - approx. Translator), following one of the most significant and noticeable trends in technology, and how it relates to any area of ​​the company itself. The main points indicated by Bezos in his letter are indicated below. This is quite interesting material for those who are interested in technology, and for those who are interested in information about the company's development strategy, aspects of the work of large corporations and the actions of the management of such companies.

Customer obsession

There are various business strategies. You can, for example, focus on the product, on competitors and the fight against them, on the business model. But the main thing is focusing on the customer, the customer. None of the customers can be one hundred percent satisfied. And even if the consumer responds that he is satisfied with the product or service, you should try to make him even happier. As an example, Bezos cites Amazon Prime, a fast and smooth delivery program. None of the Amazon customers asked to create it, but it was developed, implemented and now it is liked by very many. The problem here is that clients themselves cannot always formulate what kind of service or service they would like to receive.
')

Substitution of concepts

Here Bezos says that as the company grows, management may become a victim of such a problem as concept substitution, when the process becomes more important than the result. And in this case, an inexperienced leader can justify the unsatisfactory result by the fact that the work process was carried out flawlessly, and it turns out that nothing depended on himself.

A more experienced leader, having received an unsatisfactory result, uses all his possibilities in order to improve the process. But still, the process of achieving results is not the main thing.

A good company and its employees understand the needs of their customers. They try to spend their time and energy to improve this ability - to understand their customers. Many companies conduct beta testing with the participation of volunteers. Here you also need to understand that when conducting pre-approbation of a product or service, the opinion of beta testers may differ from the opinion of consumers.

Bezos believes that the creator of a product or service should understand the needs of customers without any research there, this skill needs to be developed. Pre-testing can help find "dead zones", see problems that were previously invisible, although they have influenced the result. But the creator of the product must have an intuition, there must be a “consumer sense”.

Follow technology trends.

According to Bezos, now the success or failure of the company depends largely on the ability of management to capture and follow global trends in technology. According to him, this is not so difficult, but for a large company, changing the principles of its work can be quite a challenge. Therefore, some companies simply refuse to implement the achievements of modern science and technology.

Bezos believes that currently the main trends are machine learning and artificial intelligence.

In the past, computers helped automate tasks that were previously performed manually. The automation process was carried out using algorithms, we needed clear rules of operation for any, even the most powerful machine. And the rules were needed even to solve the simplest tasks. Now this condition is not so tough - machine learning allows computers themselves to find solutions to some problems, and quite quickly and efficiently.

As for Amazon, machine learning has been used for several years. The results are quite noticeable. For example, these include drones from Prime Air, Amazon Go with its machine vision and Alexa, a digital assistant. The smart column Echo, whose work is based on interaction with Alexa, is so popular that Amazon does not have time to replenish stocks of this device. Nevertheless, Bezos promised to solve this problem.

For the most part, the results of using machine learning within the company go unnoticed. Nevertheless, the ranking of products and recommendations, the detection of fraudulent actions of cybercriminals, financial transactions and much more depends on machine learning. Thanks to machine learning, the company's customers receive ever better service.



Bezos argues that machine learning and a weak form of AI will continue to evolve in such a way as to remain imperceptible, affecting a huge number of operations.

Machine learning technologies themselves are also sold, so to speak. For example, “ready-to-use” machine learning frameworks running on P2 instances are used by Amazon clients for their own needs. And this is a forecast of the yield of a certain region, the prediction of climate change, the determination of the rate of spread of any diseases. Among the most well-known AI-services of the company are Amazon Lex (based on Alexa), Amazon Polly and Amazon Rekognition. These services are able to understand speech, they "speak" themselves, and also analyze images. Access to them is easily done by API. Bezos promised many new things in this direction, so we are waiting.

Quick decision making

According to the head of Amazon, companies of tomorrow should be able to quickly make a decision. In the past, and even now, it is important, but not critical (although in terms of trading in securities and other things, it is very important). But in the future, almost any important decision will need to be able to make very quickly. Speed ​​is of great importance in business.

Bezos gives some tips on how to learn and succeed in making quick decisions. In some cases, the decision can be reversed if something goes wrong. Therefore, you should not be too afraid of the wrong decision.



The head of Amazon believes that the decision should be made when there is about 70% of the information on the current issue. If we continue to collect information, considering that we need to get not 70%, but 90% of the data, and only then make a decision, then we can be late. Even if the solution was not entirely correct, then you need to learn to correct the problem already in the process of implementation.

And if the team offers something with which the manager does not agree, then it is still “not to agree, but to approve.” So you have to do it yourself Bezos, and quite often.