How a solo creator scaled a daily game to 1,000 users without writing a single line of code
- A solo creator launched a daily riddle game, The Sphinx Riddle, without writing a single line of code.
- AI agents handled core development tasks, from back-end logic to UX tweaks.
- TikTok carousel ads powered organic growth to 1,000 unique users at zero ad spend.
As a lifelong gamer, I’ve always wanted to build something of my own. When Wordle and other daily game formats exploded, I spotted an untouched niche.
Most daily word games are bland, offering little in the way of narrative or thematic flair. I set out to build something that included both my love of story-driven games and the addictive gameplay loop of daily games like Wordle.
With that opportunity in mind, I set out to see how far I could push modern AI tools - and my limited coding and marketing knowledge - to create something compelling.
The result is The Sphinx Riddle: a daily riddle game hosted by a mysterious (and occasionally menacing) Sphinx.
I built the game without writing a single line of code and scaled it to 1,000 unique users without spending a cent on marketing. Here’s exactly how I did it.
Coding - Initial build
Cursor served as my primary coding agent. After feeding it detailed instructions - game rules, visual layout, desired buttons - it quickly produced a functional prototype.
That prototype worked, but it wasn’t pretty. Since then I’ve found visual‑first builders like Lovable or Bolt far more effective for front‑end polish, but those tools struggle with the complex back-end logic that Cursor handles with ease.
Back-end logic
Cursor also powered the back-end, easily building systems that tracked user guesses and served messages from the Sphinx in certain situations. Iterating on rules and wiring up a hint mechanic with the OpenAI API was also straightforward.
I built the game without writing a single line of code and scaled it to 1,000 unique users without spending a cent on marketing.
However, the real challenge for building the AI-enabled hint mechanic was prompt‑tuning. Getting the model to give a helpful hint without spoiling the answer was very difficult and required intensive testing through a variety of prompts.
I started with GPT-4o-mini for its price, but the frequent confusing or unrelated hints pushed me to GPT-4.1-mini. The switch cost pennies (about $0.15 total so far) and dramatically improved performance.
Front‑end headaches
Polishing the front‑end through Cursor was painful. Broad prompts like “improve the UX” often made things worse.
Instead, I had success with short, surgical prompts that referenced specific HTML elements and numeric spacing I pulled while using the Inspect tool in Chrome. Next time, I’ll prototype the UI in Lovable or Bolt, then hand the files back to Cursor for back-end wiring.
Marketing: Hook pictures
TikTok picture carousels became my top acquisition channel. I paired a striking, AI‑generated Sphinx image with a caption like “Insanely hard riddle” followed by a screenshot of that day’s puzzle.
To drive further growth, I relied on data from TikTok analytics to identify top-performing content.
To drive further growth, I relied on data from TikTok analytics to identify top-performing content. I fed my top-performers back to ChatGPT, asked it to spot patterns (colour palette, brightness, thematic elements), and then generated new prompts and images in that style.
Videos
I experimented with OpenAI Sora to create fantasy clips, but stitching a polished video proved time-consuming and the results underperformed on TikTok. For now, static image carousels win.
Best places to market ranked
TikTok – King of free impressions: TikTok reliably delivers 3K impressions per carousel, with the occasional 20K spike. Finding the right niche and posting time made this a sustainable free funnel.
Discord & Reddit – Feedback frenzy: Organic posts on Reddit and Discord rarely went viral, but the feedback from brutally honest anonymous users became invaluable as I improved the game.
X/Facebook/Instagram/YouTube – Tough nuts to crack: Organic reach on X, Facebook, Instagram, and YouTube was negligible. These platforms may shine with paid spend, but for free traffic, I needed to focus elsewhere.
Conclusions
With AI tools improving daily, there’s never been a better time to publish and scale an app solo.
If you’ve dreamed of shipping a game or app, dive in! The barrier has never been lower.