A great starting point cannot be planned.

Text by Wu Bingjian

Over a decade ago, there were probably thousands of business plans (BPs) aiming to create a content - sharing community, a fashion - styling community, or a Chinese version of Instagram. Later, it was Xiaohongshu that achieved this. Its starting point was an overseas shopping guide.

Many BPs proposed creating a Taobao for the lower - tier markets to solve the rural shopping problem. Some planned to set up terminals, while others intended to use distribution models. Later, it was Pinduoduo that realized this. Its starting point was group - buying fruits.

Thousands of BPs stated their intention to build a mobile interactive entertainment platform. Some focused on games, while others on social networking. Later, it was Douyin that achieved this. Douyin started as a platform for young people's short music videos, and ByteDance's origin was Neihan Duanzi.

What does this indicate? The end - goal is often anticipated, and many can envision it. However, it's difficult to predict the starting point and the exact product form at the end. Greatness cannot be planned. Therefore, one can only do things that are valuable to users and can retain them.

In today's AI entrepreneurship, many BPs mention the next - generation AI Douyin or the next - generation Agent platform. A decade later, we'll find that these platforms do exist, but their entry points will once again exceed our expectations.

When thousands of people write about the next - generation AI platform in their BPs, it represents a common expectation, and there's a high probability that such a platform will emerge. The question is, what will this product look like on day one? What's the path from the starting point to the end - goal? This cannot be planned.

(Entrepreneurship is like...)

The common feature of the excellent products of the previous generation is that their starting points all addressed a real need.

As entrepreneurs, the only thing we can plan is - on day one, which niche group of users am I targeting, and what specific problem am I solving? Does my product resonate with users in terms of experience?

Even if the product is a bit rough, as long as the approach is right and it resonates with users, it will definitely bring user retention.

This group of users is like a laboratory. All subsequent iterations and generalizations are experiments based on this group of users. If a killer feature is launched, there's a high probability of word - of - mouth spread and organic traffic. Conversely, if the features are inappropriate, even with repeated refinement, growth will be sluggish. Users vote with their retention and payments.

This is about groping forward with the guidance of users. It's precisely this positive and negative feedback that forms the path from the starting point to the end - goal.

The process of exploring the path is also an evolutionary process. Looking back, the power of evolution is astonishing. The aforementioned Xiaohongshu and Pinduoduo not only became the so - called "content - sharing community" and "Taobao for the lower - tier markets" but far exceeded these positions and absorbed other vertical communities and vertical e - commerce platforms, growing into larger entities.

Many people questioned whether Xiaohongshu would only appeal to wealthy and fashionable women, and whether Pinduoduo would only be popular in the lower - tier markets. Now, the answers are obvious. These are products with hundreds of millions of daily active users and have generalized to the general public. Behind this generalization, each platform has its own weapon: the content quality of Xiaohongshu, the low prices of Pinduoduo, and the interesting content of Douyin. Instead of being scattered across N vertical apps, users' diverse needs have converged into a dozen or so leading apps.

The contrast between the initial vertical focus and the later generalization is more dramatic.

The initial vertical focus also applies to current AI entrepreneurship. If we aim to achieve an ARR of $10 million in the first year after the product launch, we don't need a large number of users. Assuming each user pays $500 per year, we only need 20,000 paying users. We can make choices regarding users, such as targeting only the European and American markets or only the Japanese and Korean markets, only early adopters with high tolerance or the general public, only professional creators or the general consumer group.

The core question is, have you reached these 20,000 users, and are they willing to pay?

In the mobile Internet era, there was a well - known failed case called Color, which is now forgotten.

In 2011, the hottest concept was SoLoMo - Social, Location, Mobile. Color was such a SoLoMo product born in Silicon Valley. Users didn't need to register. They could open the app to take photos, and the content would be automatically shared with users within a 30 - meter radius. It featured an "elastic social network" that automatically grouped users based on environmental data such as light, distance, and sound for social interaction. (Didn't understand it? I didn't either when I used it.)

Color's team was extremely impressive. Its CEO, Bill, was a serial entrepreneur who had founded three Internet companies, one of which went public and two were acquired, with a total exit value of $900 million. Co - founder Peter was the founder of the financial service tool BillShrink, which was acquired by MasterCard. The product leader, DJ, was the chief scientist at LinkedIn. In the angel round, before the product was even launched, it received an investment of $41 million from Sequoia Capital US, Bain Capital, and Silicon Valley Bank.

With the hottest concept, a luxurious team, and a large - scale financing, it was the star in Silicon Valley at that time. Before its release, it was highly praised by the media and hailed as a "revolutionary social app", "one of the one or two companies in Silicon Valley that change the world every decade", and "the disruptor of Facebook".

In March 2011, on the day of its release, Color topped the App Store. Within 24 hours of its launch, hundreds of articles were dedicated to discussing it. However, its reputation quickly collapsed - it crashed frequently, had a confusing operation interface, and there was little content to view. The most common comment was "useless", and 60% of the ratings were only one star. In November 2012, Color admitted failure, shut down its service, and was acquired by Apple at a valuation of $7 million.

Obviously, Color failed to resonate with any users. Great ambitions still need a reliable product to support them.

Not only is it difficult to plan the starting point of entrepreneurship, but it's also hard to plan the starting point of investment.

A common sense: If you invest in the angel round of a project and the project succeeds, when you look back at the investment memo, the product and business model will have changed beyond recognition. Only the part about the people remains the same.

Two recent widely - discussed cases, Pop Mart with a market value of $320 billion and Insta360 with a market value of $70 billion, are both like this. Their day - one states were different from their states at the time of listing. In 2012, when Mai Gang invested in the angel round of Pop Mart, it was still a trendy grocery store selling stationery, toys, and accessories. It wasn't until 2016 that Pop Mart transformed into a trendy toy company and launched its first in - house IP, the "Molly Constellation Series" blind boxes. In 2015, when IDG invested in the angel round of Insta360, the company was called "Lanfeng Chuangshi" and was engaged in live - streaming software. Later, it transformed into a company that produces 360 - degree panoramic cameras. Today, its positioning is as a provider of intelligent imaging devices.

I believe that everyone involved in early - stage investment has a similar feeling. Respecting uncertainty and evolution means respecting common sense.

A tree can grow tall not because it has many cells, but because its individual cells are strong enough. This cell is the founder and the Product - Market Fit (PMF) at the beginning of the company's establishment.

Companies that address real needs and solve specific problems on day one tend to grow taller. This applies regardless of whether it's the mobile era or the AI era.

May you become a super cell.

Author Introduction: Wu Bingjian, a partner at Soul Capital, is engaged in AI - related venture capital. He was previously a mobile product manager and strategic analyst at a large company and has worked at Xianfeng and Lenovo Star. He has participated in the investment of multiple large - model and AI application projects. Keywords: LLM, AI Native, AI infrastructure, Robotics.

This article is from the WeChat official account "AI Voyage". Author: Wu Bingjian. Republished by 36Kr with permission.