Pocket Shopping Wang: The feature of fragmentation on mobile phone shopping

Wang, the founder of Pocket shopping, is one of the people I've seen most deeply about mobile electric. Born in 1984 he is a technical otaku, but always in the thinking of the user every day in the end what to do on the phone, users generally buy things on the phone to turn to the first few pages do not want to go down these issues. He started the pocket shopping, just recently won the fund and Jingwei China 12 million U.S. dollars in a round of investment, set up only 2 years to achieve such a high valuation, a time in the creation circle caused a small discussion.

A few days ago, I had an in-depth conversation with Wang, and we focused on some of the behavioral habits of users who shopped on mobile terminals. Because the screen is so small that users are no longer likely to browse through a page of goods as they would on a PC, the user's shopping on the phone presents many new features. Because the content of the conversation is too long, I will share the main points of the conversation two times.

Zeng Ai: What's the difference between a user's habit of buying something on a mobile phone and a PC for so long?

Wang: The mobile phone shopping has the characteristics of fragmentation, after all, the screen is very small. We also do the same on the PC, but only to make a comparison with the phone. We found a very interesting phenomenon, that is, our users on the mobile terminal to collect a product average frequency is 49 times times the PC. At first I was surprised by the data, and then I figured out that the truth was simple. Big PC screen, users can look at a page of dozens of items, they can page down a page, until they find the goods they want, they see the likes of goods directly added to the shopping cart is good, do not need to collect. On a PC, people can open a lot of shopping pages in a browser on a page, and don't have to turn it off for a day.

and the cell phone screen is very small, if one to look very tired. Maybe when you turn down to page 50th, the phone is dead. So users often see a reluctant to meet their own purchase characteristics, the first collection. For example, users want to buy a razor, he saw on the phone may not be the last one he wants to buy, but reluctantly satisfied, the first collection. This gave me a great inspiration. We (pocket Shopping) now daily related sales are 400 to 10 million yuan a day, and the number of items we add each day is 400 million RMB. I was thinking that if we converted 1% of our daily collections into sales, it would be a staggering number.

Zeng Ai: How do those users ' collections on mobile phones turn into giant sales?

Wang: Last January, we found that there were more than 1000 users in the pocket shop who collected the windbreaker, but didn't buy it. Why didn't you buy it? I think the phone screen is small, not find their most satisfactory windbreaker. And then it was January, and it was cold, and the average person was going to buy a windbreaker by March. Later I thought of an idea, I find a group of Taobao windbreaker Sellers, and they said, I have 1000 people want to buy a windbreaker, you can do a group purchase? 1000 people, 1% of the people bought is 10, 10% people bought is 100 pieces.

Later we found the best seller, gave the price of 70 percent packets of mail, specifically for the 1000 people engaged in a group purchase, that is, after today is not the price, we push the activity to these 1000 people, the result sold more than 300 windbreaker. Electricity quotient general conversion rate only 1% to 2%, and our this activity conversion rate incredibly achieves 30%, the effect is very good.

Zeng Ai: After this attempt, did you find some new features of the mobile electric quotient?

Wang: I think there will be great changes in the future e-business model. In the past business model, we first produced the goods, and then went to the market to find the buyers who need it, the middle cost is very high. And now we find that users have a lot of deposits on their phones, which makes it clearer what users want. So we can sell the merchandise to people who really like it, so the conversion rate is much higher. I call this pattern "reverse group buying".

For a long time, PCs have been trying to replace television as the highest-market-share terminal, but they haven't done it, and the phone has done it in less than 10. The Internet Plus mobile phones will create the world's most disruptive business model. So we now give up the PC and focus on pure movement. We do reverse group buying, found that sellers hate the sale as a price reduction. This is not successful on the PC because there are very few people who collect goods on the PC. On the phone, less is more. The net has 1.6 billion kinds of goods, perhaps you need only 100 kinds. All we have to do is to help you pick out the 100 items. This needs to be done in the form of large data.

Zeng Ai: How did you come up with the idea of a pocket shopping thing?

Wang: When I first started this company a few years ago, I found that there was a strange phenomenon in the electric business industry, that is, no matter the size of the business, can not make money. Bi Sheng once said, oneself into the wrong line. We interviewed a lot of Taobao sellers, said they lose money. It's strange that everyone says the industry has a good outlook, but nobody can make any money. I went to see the results of Macaulay, it does mail order business, access to the cost of users accounted for 12% of the sales, physical store is 25%, and to the electricity quotient, into 35%. We interviewed most of the sellers, the cost is about 30%, so they can not make money.

I have been thinking, what caused the flow so expensive? The internet is theoretically a massive supply that is highly efficient in matching huge demands, so there is no hard business. The situation is so abnormal now. It is difficult for users to find satisfied goods, and the cost of acquiring users is very high.

Zeng Ai: What do you think is the problem of not making any money from the electricity dealers?

Wang: Later I found that users online to buy things, the most search is the category of times, such as users want to buy a windbreaker, he will search "windbreaker" two words, and will not go to search "British large-collar trench windbreaker." We assume that there are 10 million users on the Internet to search for windbreaker, Taobao to the 10 million people to provide the search results are the same. A page shows 20 results, 50 pages shows 1000. The average person will not turn down when he turns to page 50. Suppose there are 100,000 kinds of windbreaker, at last people saw only 1000 kinds of windbreaker. It is said that the current search does not study personality.

Like windbreaker such categories of keywords are not many, but also the million. 90% of the content on the internet has never been shown, not even once. Conversely, it is difficult for users to find what they like. We assume that Taobao is a 400 million cubic metre mega mall with 1 billion products. The user sees only 5% of the goods in front. The through train can only see the front 5% of the place. The 5% sellers in front do not make money because the traffic cost is too high. 95% of the sellers in the back don't make money because they don't have enough traffic. The result is that the size of the sellers does not make money.

Zeng Ai: This is a long-standing problem, how to solve it?

Wang: My first thought was who had done such a thing. Directly to find, to do this thing again, how to the goods in accordance with the preferences of users to reorder, it is good, the goods are also displayed, users stroll up also cool. We think that the user clicks on the collection to buy the record, the user has looked 20 goods, has ordered one not to point the other 19, certainly has the original. Click and purchase records are the user's different degree of purchase preferences. We use this to measure, why like, why not like.

Baidu and Google have done such things, Baidu's previous advertising system is called relevance ads, you search flowers, will appear flowers keyword ads, sorted by bid. The problem is that it has 1000 flower ads in its library, and the last one shows 10, so 50 ads are displayed throughout the four months, and 950 have not been shown. Li himself also said that historically relevant ads, 95% of the ads have never been seen by anyone. I think there is a similar situation in the field of electrical business, the whole network of physical goods, 95% of the goods have not been shown, not once.

So how does Baidu solve this problem? Later, you search for flowers in Baidu, so and the characteristics of flowers matching ads, such as gift ads, will also show. For example, if you search for a keyword that contains flowers, there may be only 1000 articles, including correlations, and 50,000.

When searching for flowers, the user may not accurately describe their purpose, but the search engine will automatically match. Each feature has its own weight, such as flowers is a gift, chocolate also has gift characteristics, with these data, through the high dimensional logic regression, calculate the weight of each feature. The result is Valentine's Day, you go to search for flowers, it is likely to appear in chocolate ads.

Baidu on the Phoenix Nest system, the number of ads displayed, from 5% to 26%. Baidu last year online personalized search technology, farmers in Baidu search for "Apple", may appear seeds and fertilizers, and technology enthusiasts search "Apple" two words, may appear the mobile phone of the Peace Board computer, the technology will show the number of ads to further increase significantly, the team also won the internal Baidu Award. We are also thinking, whether shopping is the same, according to the user's preferences, characteristics to automatically match the goods? Each product has tens of thousands of characteristics. Each of our products has oar characteristics. We rely on the past to click on the collection of search records, according to the whole network of goods according to your preferences sorted, I hope that 50% of the merchandise show you.

Search engine, turn to 50 pages, the more backward back to the weaker. If the search engine lets the user turn to the back 50 pages, it certainly dies, and the shopping area everybody generally accepts, the first few pages often can not find oneself likes the commodity, generally turns to the back many pages. We call it the discovery engine. The whole network of goods in accordance with your preferences, the order you will point, will be the collection of things in the front, according to your behavior habits personalized to recommend products to you, this greatly saves you on the phone to pick the goods time.