Image source: Generated by Unbounded AI

Investors and entrepreneurs always love and kill each other. So what happens when investors turn around and start a business?

In the past two years since the emergence of large-scale models, many investors have been unable to sit still. According to the Silicon Valley Technology Review (SVTR) AI database, among the more than 3,000 high-potential AI startups in the world, the founders and CEOs of more than 300 companies have investment backgrounds. Iconic examples are former YC President and OpenAI CEO Sam Altman and Sinovation Ventures Kaifu Lee, Chairman of the Board and CEO of Lingyiwu.com.

Li Yitong is also one of them. In 2023, she was still evaluating AIGC projects for clients at an FA. In the blink of an eye, at the beginning of 2024, she formed a team to develop her first product-Wuyuan, a second-hand trading platform based on AI technology.

However, the investment background did not make it easier for Wuyuan to obtain investment.

"Half a year ago, investors thought that second-hand e-commerce was very mature and had no new demand, so they didn't believe in us."

Li Yitong and his team had to operate at extremely low costs. After finally holding on until the product was about to go online, new challenges came again.

Like all AI entrepreneurs, she hoped to avoid the most competitive track and quietly enter the market through second-hand e-commerce. However, Wuyuan’s head-on encounter with the giants came faster than expected.

On July 26, Xianyu announced that it would charge a 0.6% service fee without distinction for collective sellers. For sellers who place an order quantity of more than 10 pieces and a cumulative transaction amount of more than 10,000 within a natural month, 1% of the transaction amount will be charged as a service fee for each order after the excess.

On July 27, Wuyuan Xiaohongshu’s followers increased by 10,000. Wuyuan officials responded to user needs and issued a statement that they would not charge basic software service fees. In August, on the first day of pre-launch of the Wuyuan WeChat mini program, tens of thousands of users poured in within 20 hours. After the WeChat mini program was officially launched in September, the transaction amount on the platform reached one million yuan on the 45th.

For a product that is still in internal testing, this traffic can be a blessing in disguise, or it can be a life-threatening talisman.

A large number of users come to compare Wuyuan and Xianyu, and what they are most concerned about is the "charging issue". But in fact, Wuyuan relied on the WeChat applet, and WeChat stipulated that it would charge a handling fee of 0.06%, which triggered a new round of doubts. When a product is just budding, excessive traffic may amplify immaturity, leading to the loss of potential users.

Amid the controversy, we chatted with Li Yitong, founder and CEO of Wuyuan, about her understanding of AI second-hand e-commerce and her views on competing with giants.

The following is a transcript of the conversation:

#01. AI + second-hand e-commerce, where is the imagination?

Silicon StarPerson: Please introduce the circle of things in a paragraph.

Li Yitong: TreLoop is an AI-driven idle item sharing and trading platform. Users can communicate with each other based on personal interests, price and sell their own items. Compared to "second-hand", we prefer to call it "idle" to achieve the perfection of the item, so it is called Wuyuan.

Silicon Star: There are many second-hand e-commerce platforms in China, such as Xianyu, Zhuanzhuan, Aihuishou, Hongbulin, etc. Why do you create a new product? What's the difference? And what is its relationship with AI?

Li Yitong: As the Internet reaches today, there is no field that is not a red ocean. From the perspective of the overall economic situation, young people’s changes in job search, employment, mental and social life, as well as their pursuit of personalized interests such as the second dimension and the three pits, will cause the second-hand market to continue to grow in the next three years. This is a rare blue ocean market. In today's second-hand market, there have been almost no new products in the past five or six years. Now that AI is here, we predict that this is a new opportunity.

In terms of products, our difference is that we attach great importance to the community. The end result is 60% transaction and 40% community. We will also infiltrate AI into various functions of the product to better satisfy users.

Silicon Star: It sounds a bit abstract. How is the AI ​​function specifically implemented? +What is the value of AI?

Li Yitong: AI is actually just an auxiliary. We first need to build an idle trading platform based on the community, and then integrate AI into users’ links to better serve users.

In this scenario, my prediction for AI e-commerce is analogous to the search scenario. Users will think that Kimi is better than ordinary Google because Kimi can not only help with search, but also with summary. , refining, essentially analyzing the intention. In an e-commerce scenario, through your dialogue with AI, for example, if your intention is to learn skiing, AI can be extended to sort out the equipment that novice skiers need, select styles, models, and cost-effective ranges, and help you You get it all in one place. In addition, in the entire e-commerce shopping link, such as when users put products on the shelves, AI will help to recognize pictures, write introductions, and in the search process. In addition to our own recommendation algorithm, we also superimpose a Layer AI model for more accurate matching.

Silicon Star: I understand. In fact, Taobao has made similar attempts. Recently, Perplexity also wants to recommend products, but the AI ​​recommendation model is difficult to combine with traditional search e-commerce. User habits, traffic distribution logic, and business models cannot be combined. Do you think this is a problem?

Li Yitong: We are not saying that we have to do AI. It is not necessary. If other second-hand platforms add AI, won’t we have nothing to do with it? Regardless of AI, the platform is still established. Moreover, AI’s current understanding of users and products still requires continuous training. Our current users have generated a lot of questions about AI when using some of our AI capabilities.The data needs to be cleaned and retrained. We are also guiding users to talk to the assistant. In fact, during the internal testing in July, ordinary users did not have a high understanding of the acceptance of AI. It is different now, and the effect and acceptance are changing.

#02. I don’t want to have sex with Xianyu, I would rather do Xiaohongshu

< /p>

Silicon Star People: As soon as you went online in August, your data skyrocketed. At that time, Xianyu started charging handling fees, and there was a lot of discussion about second-hand e-commerce. Did it affect you?

Li Yitong: When Xianyu charged fees, users began to list idle platforms on the market, and we were included in the discussion. For us, it was considered event marketing. However, before that, we also made a lot of preparations in advance, updated many social accounts, and asked internal beta users to write evaluation content, etc., in order to get traffic in the short term.

Silicon Star: In this wave of AI entrepreneurship, everyone is afraid of facing off against big companies. Will you be panicked when you go up against Xianyu as soon as you go online?

Li Yitong: Actually, we are quite different. Before we made Wuyuan, we collected complaints from users on basically all second-hand platforms in China and truly saw the needs of users, so we created Wuyuan. My anxiety point is that after the traffic comes, we have high expectations.

Silicon Star: What are the specific demand points?

Li Yitong: Currently, the most idle users in China are the post-00s generation. They will have great value and will grow up and become consumers in the next two to three years. Main force. They have extended many methods of idleness, such as bartering for things. Some idle unboxing bloggers unbox a replacement item every day, often with small cards, snacks, etc. given by the other party. This is highly interest-oriented. They are both price-sensitive and interested in non-essential and non-standard products. Interested and more focused on the emotional and spiritual level.

Silicon Star: So Wuyuan is more like an idle sharing and communication community. This is the difference between you and other traditional second-hand platforms.

Li Yitong: Yes, in fact, we did not position Wuyuan as a pure trading tool or platform from the beginning. Unlike other platforms that are purely item-based transactions, Wuyuan wanted to create a New way of life.

Silicon Star People: Rather than Xianyu, you want to be the Little Red Book in the world of idle items.

Li Yitong: Yes, I have also mentioned this to investors.

Silicon Star: However, second-hand platforms are almost all heavy transactions. Xianyu also paid attention to the community in the past, such as promoting transactions with communities, fish ponds, and seafood markets, but in recent years it has also been transitioning to transactions. Coupled with C2B2C models such as Zhuanzhuan, Aihuishou, and Hongblin, have you ever wondered why no one cares about C2C? Can C2C idle transactions run smoothly in terms of security and business model?

Li Yitong: Once the C2C platform model is established,The network effect will naturally generate benefits, and there will be many ways to develop business models in the future. We don’t do C2C. In fact, we don’t do the platform model now. We feel that it is difficult to achieve good results in investment, and it is difficult to achieve large-scale profits at the same time. But coming back to our products, we didn’t spend money on advertising. Users will naturally look for new platforms to meet new needs and then find us. Even in terms of total cost, we did not spend too much money. This is different from the previous era of money-burning wars on the Internet. In the past, we were really attracting new users and educating the market, but now we are not.

Silicon Star: Wuyuan promised not to charge handling fees before. What is the business model?

Li Yitong: We currently have two main commercialization ideas: one is membership value-added services, and the other is brand supplier cooperation. In terms of membership value-added services, we can provide users with some paid value-added functions. In terms of brand cooperation, we are in contact with some brand owners, and we will charge a certain commission if we get some of their inventory, tail goods, etc.

Silicon Star: Let’s continue talking about transactions. Transactions require a security system. Is it difficult for startups?

Li Yitong: We did encounter a lot of problems, but we also eliminated a lot of competition. We have spent a lot of energy on transaction protection. For example, when we first launched the mini program, WeChat required users to upload their ID photos, and many users were very resistant to this. Many problems need to be solved step by step.

Silicon Star: Transactions must be based on a community atmosphere. In terms of transaction + community, Xianyu has transformed, and Xiaohongshu is also struggling to balance. How do you plan to build a community, and how do you balance community and transactions?

Li Yitong: Community building is indeed a very challenging thing for us. First, content is the foundation of a community. We will set up corresponding content sections around some subject areas that young people are interested in, such as two-dimensional, handicrafts, trendy toys, etc., and encourage users to share their collections, creations, etc. here. Nowadays, many young people have their own interest circles, which has led to the trading of hairpins, dolls, figures and other items.

Furthermore, in terms of product design, we think about users’ usage habits in transaction scenarios. We turn the community into a two-column picture, text, and video information flow on a social platform. We also pay attention to the construction of personal accounts and display the user's MBTI, interest style, etc. on the homepage. Our user's self-introduction has a column for zodiac sign and MBTI. Based on this, AI will analyze the user's browsing habits and recommend users and products that are more similar to the user's. For example, if an ENFP and an INFP exchange items, 6 users and 12 other items will be recommended to the user in the form of drift bottles every day to help him find interesting items or friends.

The emphasis on community is reflected in many places. For example, as a typical example, on other second-hand platforms, buyers and sellers have conversations based on "items". I like three of the other party's products, and I There are three dialog boxes to be launched, but in the object circle, there will only be one dialog box. This is the sameGood "people-to-people" dialogue.

#03. Investors can’t get investment when they start a business?

Silicon Star: After talking about so many products, please introduce yourself.

Li Yitong: I used to do product planning, operations and integrated marketing in Internet companies such as iQiyi and Tencent. He has two entrepreneurial experiences and has held the roles of joint venture and CMO. Last year, I mainly watched the incubation of AIGC early-stage entrepreneurial projects at FA, and this year I decided to start my own business again.

Silicon Star: Did your experience at FA help you start your own business? Ordinary entrepreneurs may determine an entrepreneurial project based on their own advantages and interests combined with market judgment. You have seen so many projects, why do you want to start an e-commerce business? Why did you decide to do it yourself instead of asking someone to vote for you?

Li Yitong: Last year, everyone felt that the tracks with great opportunities were overseas or AI. I first identified the opportunities for this wave of AI, and then applied this layer to the large model, excluding to B, because there was no corresponding With the resources and business development energy, I chose To C because the market is relatively fair.

Secondly, e-commerce is actually a market with a large market but far away from AI. I think we have time to get ahead and explore. On this basis, we have selected more subdivided second-hand idle markets. This is a continuous search process. As for why I don’t directly invest in other people’s projects, I have always wanted to start my own business, and I also invest in order to eventually accomplish something big by myself. The fundamental point is that as an investor, it is difficult to have a high degree of participation and influence on the project. I hope to delve into this field personally and realize my ideas thoroughly.

Silicon Star: What do you think is the difference between investing and starting a business? Investing should be a lot more fun, right?

Li Yitong: Everyone has different hardships. Investment and entrepreneurship are actually people in the same boat. The money invested cannot be invested, or the investment was not done well last year. It is difficult to explain to the LPs, and the management fees cannot be returned, which is also difficult. Now everyone seems to be complaining that the industry is going to collapse.

Silicon Star: What difference has your investment experience brought to your entrepreneurship? Will it be easier for you to understand the market? Or will it be easier to get investment if you better understand the needs of investors?

Li Yitong: I personally feel that after watching many projects, my judgment on the industry may be stronger. I will be more objective in assessing where the opportunities and challenges of a project lie. At the same time, having experienced the investment process has also given me a clearer understanding of what investors are considering and paying attention to.

Silicon Star: But you don’t seem to have received the investment.

Li Yitong: Yes. This was something I didn't expect. I originally expected a sum of funds to come in in July, but it didn't arrive due to various reasons. But I didn’t expect that even without external fundingThanks to your support, we have reached this point. We only spent a small amount of money and achieved a transaction volume of 1 million in 45 days. Some investors told me that we were already successful. In the past, we could have raised a lot of money.

And there is not that much consensus among investors. When looking for investment, you need to find people who understand each other and have consensus. Many people don't believe that the platform model can still be built now, and are not interested in second-hand e-commerce. They feel that this market has solidified, and they are unwilling to invest.

During the last round of chatting, an investor told me that if users don’t have the needs I have now, we can’t do it.

Silicon Star: Do they believe it now?

Li Yitong: I believe it now.

Silicon Star: Do you still want to take the capital route in the future, or will you rely on user income to make a positive cycle?

Li Yitong: Depending on the expectations of things, of course I hope to take the capital route, because that means doing something bigger.

Silicon Star: What are you going to do if you never get investment?

Li Yitong: Everyone has their own approach. I will also make a new product next. One model may be regarded as luck, but two or even many models can still come out. No matter how good our team’s judgment is, the product capabilities and error correction capabilities can be verified, which may make investors optimistic about our team. Maybe we It is possible to rely on new products to get more user payments.

Moreover, the development of the first product has been basically completed, the team has matured, and some development modules have been developed. It will be faster and easier to develop the second product. Everyone can look forward to it.