My Next Steps as an Independent Developer
Do not chase trends. Pursue the question you will still want to answer three years from now.
After about six months of practice and learning, I reviewed where I stood and considered what to do next to progress as an independent developer. I arrived at the following directions.
Build products that can be sold
The idea is to put what I have learned into practice: package professional skills into products that can be sold and delivered.

- Build small, focused products. I was inspired by take-a-screenshot.org, which I had seen ranking first for “how to take a screenshot.”
- Put 100 sites on the agenda. Inspired by a video creator, I want to use launching 100 sites as a concrete way to counter too much thinking and too little practice. The number could also be 500 or 1,000. Why do I consider this worthwhile? During a project, I often ask what the effort is for. Without positive feedback, it is easy to stop. A concrete target could help me avoid giving up or standing still.
- Approach early work with curiosity. A veteran Reddit user's perspective resonated with me: start out of curiosity rather than making money the immediate purpose. Otherwise, difficulties can easily make me retreat.
The line I recorded was:
Building to learn is way more valuable than building to get rich, at least early on.
My implementation plan is to find needs in Reddit posts, then use Affordablekeywords to check related keywords, their difficulty and search volume. At this stage, my principle is to build when I find a real need. Over time, I expect to narrow the focus toward opportunities where ranking looks more attainable.
If one web page can solve the problem, I will launch a small subsite. If it requires a more complex solution, I will first write a blog post about it. I plan to connect Google Analytics and Google Search Console, look at the data a week after launch, and decide what to improve. One week is an arbitrary review interval; if there is not enough data, I will keep observing.
Once there is some user data, I can use it to decide the direction of the site, such as expanding to new keywords or adding features.
At a later stage, I see two monetization routes:
- Advertising: apply to Google AdSense or Adsterra and monetize traffic. My original notes describe Adsterra as having a low application barrier and approval within one working day, while AdSense was becoming harder to join.
- Paid access: choose monthly or annual subscriptions, or credits that do not expire, depending on the product.
By repeating this process and launching 100 sites, I hope to develop new insights and move beyond the beginner stage.
Reddit is one of my preferred starting points for finding needs, but there are two other approaches I want to use:
- Start with my own problems. If I need a time-management tool and cannot find one that fits, build it, connect it with relevant keywords, and follow the same process.
- Explore new terms in Google Trends. Investigate queries with rising interest and launch quickly when the analysis suggests an opportunity.
Combine breadth and depth
How should I approach independent product development? Drawing on roughly six months of practice, lessons from experienced indie hackers and my existing projects, I have three directions:
- Depth A: continue promoting products whose process has already been validated, through backlinks and social media, while analyzing user behavior.
- Depth B: use existing users and keyword information to decide which features to add to launched products.
- Breadth: explore new needs through Reddit, my own experience and emerging queries.
Together, these form the basis of my plan.
Choosing an AI coding agent
My conclusion at the time was to switch from Claude to Codex or Grok.
Repeated Claude account restrictions had disrupted my work, including when I used a remote VPS. Each interruption sent me looking for a backup. After trying Codex, I found its capabilities strong enough that I decided to use it in place of Claude for the time being.
I also borrowed access to a Grok account and found it useful. Both fit my local workflow without requiring me to log in to a remote VPS, so I was considering which of the two to choose.
Grok interface

Codex interface

These are ways of putting the work into practice. Maintaining the right mindset matters even more to me. The Reddit screenshot below expressed that perspective well and felt like a fitting conclusion to my notes.

Editor's note, October 24, 2026: this translation preserves the original plan and its monetization options. The 100-site target is a plan, not a completed result. The screenshot ranking, advertising approval time and AI account experience describe the author's notes at the time; the ranking date, approval policy and screenshot capture dates have not been independently verified for this update.