Amoy Search: A 0 to 100 million of cock silk entrepreneurial reverse attack

Wen/Jinguo dao (micro-credit public number Ijincuodao) I am very concerned about the image strategy of the product, previously done similar. Recently, a friend of the investment has recommended a picture shopping search case Company, 3 years, from 0 to achieve 100 million users, quite cock silk reverse attack. Image search is a difficult to do the market, on the one hand directly face the BTA picture strategy, on the one hand to face the beauty of such as shopping Guide website competition pressure. With Amoy Search founder Lin Jiancong chatted, once took Taobao first win in Taobao application developers Contest Championship, but really become products, also walked a detour. Starting in April 2010, a detour was an early build of the search architecture and technology, but without users, there was no understanding of the user. Another detour is that technology is easy to go to the dead end, how to technology supremacy. This product now accumulates 100 million users, and millions of users are using it every day. 15 million times a day to process a picture search request. The biggest pain point is: To solve the user can not and do not want to use text search difficulties! My judgment, this product wants to be big, still have two big challenges: 1, want to spell Daddy, Alibaba invests this company, but want to be in the picture shopping search on the product threshold of the competition, need Ali in the data of greater open. 2, to spell service, the current Amoy search more than 100 million of the users are mostly from the "Amoy Shopping assistant", the tool is free, service to make money, how to spell outside the tool service, is a big problem. Golden wrong knife: Robin Li once said that the new reading era has come, "China's Internet content mainstream, is emerging from the text to the picture trend." What do you think of the biggest pain points? Lin Jiancong: We found that users have a lot of scenes, see the likes of things, but because it is not clear enough to use words to express, can not go shopping site search. Use words to express a see things, such as a clothing, the brain translated into keywords, in fact, is a very difficult thing for users. So, we see the user search on Taobao keyword, most of them are vague keywords, such as: Korean version, 2013 new style, dress, few people on the product style and design elements such as real decision shopping key features can be clearly described, especially in the face of a clear want to buy goods, it is difficult to describe, search in the shopping field, Very typical is the clothing and other non-standard class, accounting for the entire E-commerce 55% market share. Now users, in Taobao as long as you see a certain commodity, you can go to Amoy search directly with pictures can be searched with the same paragraph and those stores sell, can also find more similar money. This is the past, text Search can't do. For example, we see a lot of questions and answers in the Baidu request, "Edison Chen this sweater where to buy." In the shopping scene, the user picture to look for the merchandise, each day our search request already more than 15 million times. In the information, portals and other sites, we and two portals have been cooperative testing, found that the flow of these sites, about half are looking at the picture, which, shopping, women-related PV are over billion. In these sites, after reading the demand is cut off, the user sees like the eastWest, want to find no place to find, demand into a break ring. Yahoo China before the transformation, we also and Yahoo to do a test, found in the picture to provide users with fast search, to find similar goods services, users use a high viscosity, or even the number of clicks are ordinary ads several times. Surprisingly, the CTR is always growing. Description, this kind of service, really useful to the user the subversion of the portal is a more personalized and accelerated information acquisition, we try to use image search to complete more accurate and faster than text search. This change, the user experience will be very obvious. In the past, said not clear, can only knock more extensive categories of keywords, search a range, and then slowly turn over many pages to find. Maybe a lot of people can not find the desired results for half a day. A true and interesting story, an old lady, her daughter taught her to go shopping on Taobao. Later, she saw a very like clothes, the shelf. How to use text search did not search, toss the old lady for several days, the mood that tangle. Her daughter is also known as Taobao senior buyers, but also toss for several hours, how to search are not found, this is now the plight of text search. Later, know Amoy search, she use QQ screenshot, upload to Amoy search, find a lot of the same clothes in different stores still sell. In the online shopping platform, between buyers and sellers, for subjective and objective reasons, the description of the same product is not the same, users can find the same paragraph is only a very small part. In addition, the seller in order not to be people parity, as well as SEO, always artificially write the title of the product, so that buyers are not found. However, the problem of image search is expected to be completely resolved. Golden wrong knife: Before there are many companies want to do personalized shopping search engine, but feel not enough pain point. Lin Jiancong: Some products want to do is recommended, strictly speaking is based on user behavior recommendations, looking forward to more than Taobao recommendation engine, to users to push more accurate merchandise. In fact, I think this is very difficult, on the one hand, it is difficult to establish users can perceive the value of the recommendation is not obvious; On the other hand, Taobao's recommendation has been very strong, which the user migration cost is higher. Text Search in Baidu, Taobao so powerful situation, we think is the Red Sea, entrepreneurial success probability is very small, we do not touch. In fact, at the beginning of our establishment, Ali invested us, we have been thinking about this issue. We try to do is the user needs homeopathic products, and Baidu, Taobao text Search has a clear line of products, can play our team expertise products, of course, the market size is large enough, so we chose the image search. From the user search source, the entrance to the business, user value and now a large company area separated. Although, image search is the world's internet problem, very difficult, however, the road behind the existing market is huge, business model mature, competition is relatively small, the main test is our patience, confidence, to grasp the needs of users, sinking down to do technology and application innovation. Image search can actually be applied to a lot of areas, we focus on the current shopping field breakthrough, the main reasonStill want to find users really rigid image search needs, but also convenient for us to do technology, application breakthroughs. We decided to hit the ground before the shopping search field, also done a lot of research, which found: more than half of the flow of Taobao, sold in the apparel and other non-standard categories, and in these categories of users search for the majority of the keyword is fuzzy keywords, users really do not know the characteristics of costumes, as described above. At the same time, we also do a large number of Taobao real shopping user EDM research, more than million user feedback, proof of image search is the majority of users rigid demand. Golden Wrong knife: This tool is very painful point. You say the number of more than 100 million users, Amoy search (Image search engine) + Amoy Search Assistant (Amoy page plugin) + Find God Horse (APP). A total of 100 million? Lin Jiancong: This is the real statistics, the user who used our products cumulative I looked at about 110 million, now retained users almost 70 million up and down. Golden wrong knife: from 1 to 100 million, tell me about your cock. The most critical product innovation point? Lin Jiancong: Image Search, the plate is very large, but technical and application breakthroughs are particularly difficult. We are doing business groping while developing technology. Probably experienced four stages: the first stage, the company was founded in April 2010, the completion of the technical framework, the establishment of Amoy search site, trying to guide users to Amoy search site to experience the picture to find goods, six months of time, found very difficult. Both research and development, but also to find a considerable number of users, and, at that time, the characteristics of the user we are lack of understanding, only feel that shopping users, user characteristics, shopping real understanding of the environment is not enough to stay, not enough to get a sample of demand. For a start-up, it's really hard to do a search and do an image search. The second stage, because we and Taobao years of cooperation, investment relations, but also because Taobao also saw the value of the application direction, so we intend to learn Google, Baidu Early development, to provide services to Taobao. To Taobao related lines of business, providing image search technology, this is a key step. Let us have a real user environment, enough users, we have a better understanding of what users need in the shopping scene. It turns out that because of the value of image search, we provide the product contrast text search, the average turnover to upgrade at least 30%, the number of individual categories to reach 200%, so, Taobao many lines of business and we cooperate, ordinary daily UV over 800,000, the peak of the time UV over 1.5 million. This gives us a lot of confidence, and then we start out and leave Taobao to hatch the soil. To get closer to the user to dig up the demand, so we started to work on the independent application of the product. The third stage, we first to Amoy search the main station revision, but a number of consecutive changes, found: The user's understanding of image search is not enough, users want to take pictures to Amoy search methods, or too cumbersome to promote users, precipitation users or too difficult. So, we think about what is the easiest way for users to use it. We think of two methods, one is the website cooperation, take the alliance way, put generationCode, the other side of the site can be a quick search of the image, the second is to provide users with Web Plug-ins, users installed in shopping, information station to see the picture can be fast search. In a word, in the picture up stall, put fast search, from the user recently. Now, in the fourth phase, we found that the demand for user mobile phone camera search exploded, so we started experimenting with apps and building the full layout of the company's PCs and phones. Golden wrong knife: from 0 to 100 million, from the product manager's perspective, what is your biggest reflection and experience? Lin Jiancong: From 0 to 100 million, from the product manager's point of view, my biggest reflection: the product must be homeopathy, not to take for granted to the user to create a product. Do not always fantasize, go to another place to find blue ocean, but should really sink down the heart, how to insert their own blue flag in the Red Sea. Wide market innovation is the easiest to succeed, bigger, because, in the Red Sea will have sufficient size, mature enough users. Our company is not the same as many technology companies, we are very clear that the whole staff seriously said: We are an application-driven technology companies, not the usual technology-driven companies. This determines that we are always, think of how to find the user rigid demand and convenient use of experience, secondly, is to conceive technology and product implementation path. It's easy to get into a dead end because it's hard to get people to play. To do technology innovation companies, often the most difficult to break through the place there are two: first, the need to really use the technical algorithm breakthrough; second, how to make limited technological breakthroughs to find the best application of soil. Market and technical team, in technology companies, all the time is fighting, we have also had a tangle. However, after the dispute, running-in, we found that the solution to the problem in addition to balancing the technology and market team voice, give a clear value positioning, the most important is the user needs, what users want, we can do.