靠AI营销产品做到3万美金月MRR

Building in AI

I have over five years of experience in the internet industry, and I’ve worked in both engineering and product roles at leading tech companies. But my love for entrepreneurship brought me to become an independent founder. I enjoy the freedom of building things I genuinely care about. Being able to work on what I truly enjoy in a way that feels both free and fulfilling is incredibly important to me. I'm currently focusing on three products. One is a product called Leadmore AI, a B2B product focused on AI marketing. Leadmore AI is currently doing over $30k MRR, and it’s still growing very quickly. Another is a B2C product centered around an AI-powered entertainment community. And I'm also building another AI marketing product focused on the GEO space. I chose to build in AI because I strongly believe AI will fundamentally change many of the assumptions behind the traditional internet and mobile internet. It’s creating a massive new wave of opportunities and an entirely new market.

A process for moving quickly from idea to product

My process can roughly be broken down into a few steps: First, you start with an idea. That idea usually comes from insights into real user needs, especially from observing conversations and behavior on social media. And ask yourself, "What am I actually good at? In which areas do I have a deeper understanding or stronger industry insight than most people?" That’s usually the space you should focus on. Second, you validate the idea. Instead of jumping straight into building a product, I prefer validating it through operations and content first. That means sharing demos or even just the idea itself on social platforms, seeing if potential users show up, and talking to them directly to check whether their real needs match your assumptions. Ideally, you also find ways to test whether they’re actually willing to pay. Validating demand before writing any code is extremely important. Third, you talk to potential users. I usually aim to reach 50 to 100 people, but even ten deep conversations can dramatically improve your understanding of the problem and help you prioritize product requirements much more clearly. Finally, you build the product. With vibe coding today, you can often ship a very basic MVP in one or two weeks. At this stage, the key is to stay truly minimal. If one feature is enough, don’t build two. Ship fast, get it into the hands of early users, and iterate quickly based on real feedback. In the AI era, this speed matters more than ever.

A credit-based model

Our business model is credit-based. Users purchase credits, which are then used for actions like posting, commenting, or discovering relevant subreddits. Any unused credits can be refunded at any time, which makes the model very user-friendly. In terms of revenue growth, the main levers are expanding our user base and improving retention. You can think about revenue with a simple formula: new users multiplied by the conversion rate multiplied by the retention rate. Among these, we care most about retention. If retention isn’t strong, it usually means the product isn’t delivering enough value. Only once retention is healthy do we focus heavily on acquisition efficiency and conversion rates.

Serverless architecture

Our entire system is deployed on a serverless architecture. We use: Next.js as a full-stack framework for the frontend, Go with Gin to power high-performance APIs, MongoDB for core business data, ClickHouse for analytics workloads, Function Compute for background and task processing.

Fight the instinct to expand

As you’re building, you'll start to notice more competitors. They might even begin reaching out to you. Many of them ship a wide range of features, and that’s when you start questioning whether you should do the same. But when your team is small — often just one or two people early on — it’s almost impossible to execute many features well or iterate effectively. Human nature pushes you to want more, but this is exactly when product building requires subtraction, not addition. You have to fight that instinct. If I could do it all over again, I would reduce my first MVP from three features down to just one. That alone would likely have allowed me to ship within two weeks, and the overall pace of progress would have been much faster than what I experienced.

Growth via content marketing and relationship building

Most of our user acquisition comes from operations and content-driven growth. In simple terms, we create and share content across social platforms like Reddit, where we talk about our industry, share practical knowledge, and explain the value behind what we’re building. When someone shows interest, we follow up with direct conversations. If they’re genuinely engaged, we invite them into our private community to continue the relationship. That’s our main approach to user acquisition. The second part is what happens after users start using the product. We stay in close, ongoing communication with them to understand their needs, pain points, and where the product still falls short. This kind of long-term user engagement and relationship building are just as important. It not only helps us refine the product, but also naturally leads to word-of-mouth growth over time.

Three things every indie dev needs to know

My advice mainly comes down to three points. First, before building anything, spend a significant amount of time on user research and talking directly to users. This could be one month, two months, or even three months. Until you truly understand the demand, don’t start building. Many indie developers default to building first, but I think that mindset is fundamentally flawed. Second, indie developers today must strengthen their operations and growth skills. If operations aren’t your strength, find a cofounder who is strong in that area. In today’s environment, operational ability can, in many cases, be more important than pure development skills, and the overall bar for indie founders is much higher. Third, indie developers should avoid blindly chasing trends. That approach is very likely to fail. What really matters is clearly understanding your own industry strengths and continuing to iterate around them. With sustained focus and accumulation in one direction, you’re much more likely to achieve meaningful results.

What's next?

My short-term goals fall into three areas. First, I want to keep scaling Leadmore AI’s revenue while continuously improving the product experience. This is a very pragmatic and immediate priority for me. Second, within the next two weeks, I plan to launch a new product focused on GEO. Our goal is to build something truly differentiated and aim to become the leading product in that space. Third, over the next year, I want to explore more innovative possibilities in consumer-facing AI and entertainment. This area comes with more uncertainty, but it also has much higher potential for creative innovation. You can follow along on X. It’s mostly in Chinese and not very actively maintained yet. You can also reach out at [email protected] for conversations, collaboration, or promotion opportunities. Anyone with Reddit marketing needs is welcome to explore leadmore.ai. And here's my new product, modelfox.ai. It's currently a very rough demo. It will be officially launched within two weeks, so interested users can apply for early access.