Chapter 1399

Li qiguang touched his hair thin head, then got up and went out to do business.

"Uncle Guang, I find you've lost a lot of hair recently." The sky looked at his head, and his hair became more and more scarce, which would soon become a desert.

Li qiguang said with a smile, "yes, it will be very smart soon!"

"Is it too much pressure at work? I'll give you a vacation," he said with a smile

"No, it's genetic. My dad lost all his hair when he was 50." Li qiguang said with a smile.

Lin Keqing said, "you can wear a wig."

"What do you recommend?" Li qiguang asked

Speaking of recommendation, Fang Tian really thought of a very important thing. He opened Qianjin mall and browsed it on the website.

It's not ideal to find the product recommendation of the website.

Looking at the frown of Fang Tian, Li qiguang felt puzzled: "what's the matter, you also have the problem of hair falling?"

Fang Tian points to the computer screen and focuses on the product recommendation on the web page. "Did you find any problems with the recommendation of the website?"

Li qiguang came over to see: "what's the problem? They are all products recommended by the website to consumers. "

"But I'm a man. The website recommends high-heeled shoes to me. Don't you think it's silly to recommend such shoes on the website?"

Li qiguang said, "it must be like this. The website doesn't know whether you are male or female."

Fang Tian shook his head: "this kind of recommendation is not what consumers want at all. What I need is personalized recommendation!"

For e-commerce shopping website, recommendation system is very important. It can guess what consumers need, and then recommend commodities to consumers, so as to improve the sales of commodities.

In this year, there is no intelligent personalized recommendation at all.

There are basically two forms of recommendation system for all websites.

First, the website editor selects manually, and then places the product in the website recommendation position.

The second is random recommendation.

Sorted by time, the latest products will be put on the recommendation.

But either way, it is not intelligent personalized recommendation.

When a consumer opens a web page, everyone sees the same recommendations.

For example, the product recommendation on the home page seen by Fangtian now.

The first is high heels.

The second is leather shoes.

The third is children's toy car

Whether it's your account, or Li qiguang's account, or Lin Keqing's account, the recommendations you see are the same.

Such a recommendation is very unreasonable.

Lin Keqing said: "I also feel that such a recommendation is unreasonable. If the website can closely guess what I need, it can improve my purchase interest."

Fang Tian orders Li qiguang to call Yang Fan, who is in charge of the website design.

Immediately, Li qiguang called Yang Fan.

Before long, a black faced young man came in. Yang Fan's skin was of this color. I didn't know that he was from Africa.

"Brother Tian, what do you want me to do?" Yang Fan pulled over the chair and sat down.

"The website needs to make a recommendation system," Fang Tian said

"Important?" Yang Fan asked.

"Very important." Fang tianyuqi affirmed: "this can greatly improve the user's activity, more importantly, it can improve the conversion rate of goods."

The conversion rate, to put it bluntly, is the volume of views and transactions. 1000 people open the recommended products page to view, and only one person buys, which is too low.

If the recommended products are of interest to consumers, the volume of transactions can be greatly increased.

Yang Fan understood that to know the personalized needs of consumers, the key is to recommend algorithms.

"This recommended algorithm, how to do it?"

He felt his chin for a while. "The first is to recommend products to consumers according to their browsing records and search records, which is the simplest and easiest way to guess the needs of consumers."

This is not hard to understand. When a user goes to an e-commerce website for shopping, he searches for something and browses something, but does not place an order immediately. The website can recommend it to the user.

Yang Fan nodded: "what about the second way?"

"Second, it is recommended according to the purchase records of consumers. We can recommend the products he has bought with high praise and this product is not durable goods. "

Yang Fan thought about it for a while, and understood it completely. Once bought to give the high praise, the proof consumer is satisfied with this commodity, the website can recommend to him.

However, it can't be durable goods. For example, consumers bought a mobile phone in the mall half a month ago, and can't recommend another one.

But if a bag of Nestle coffee, well, this can be recommended to consumers again.One day, he took a sip of coffee and said, "the third one is recommended according to the surrounding products. For example, three days ago, when consumers bought a mobile phone in the mall, we can recommend products related to mobile phones, such as a headset or a protective case. "

"For example, if a man buys diapers, we can recommend him to buy beer."

Yang Fan wondered, "is diaper related to beer? The man who buys diapers will buy beer? "

"You have to ask Li qiguang about this."

Li qiguang's son is just born. He has the most say.

Li qiguang said with a helpless smile, "it's a good recommendation. Every time I buy diapers, I feel very depressed!"

"Ha ha!" Fang Tian and Yang Fan laughed.

"The fourth recommendation method is classified by users. The website classifies consumers according to their age, purchasing power, interests and hobbies. "

"With this intelligent algorithm, we find that a and B are the same kind of consumers."

"If a customer buys a comic book and then an Altman toy, then when b customer buys the same comic book, he can refer to a user and recommend an Altman to him."

Yang Fan, Lin Keqing and Li qiguang listened to Fang Tian's recommendation algorithm and felt very interesting. They would be more intimate if they did what he said.

In the following time, Fang Tian talked a lot, and Yang Fan recorded it.

This recommendation system is simple to say, but it involves very complex intelligent algorithms and technical matters, which need to be optimized and explored by technicians.

With this powerful recommendation system, the shopping experience of the mall is far away from other similar websites.

Finally, Fang Tian said, "this recommendation system can be used not only in shopping websites, but also in other soft cloud applications."

Soft cloud doesn't just have e-commerce websites.

Its soft cloud news, if this algorithm is used, the recommended news will be closer to the user's interest.

There are also soft cloud video, soft cloud microblog, Jinyu novel network, etc., all of which can use this system.