I have a graveyard of dead side projects. If you’re a developer, you probably do too.
There’s aqt-bs-bot — a Telegram bot I built, used twice, and abandoned. There’s gooyabot, another bot that went absolutely nowhere. There’s 처치테이블, a project whose name I can barely remember the purpose of. And scattered across my GitHub are half-finished repos that represent hundreds of hours of work and exactly zero dollars of revenue.
Sound familiar?
As a Korean developer building things in my spare time, I’ve lived the classic developer side project cycle: get excited → build something technically cool → lose interest → never monetize → repeat. For years.
But on May 28, 2026, I started something different. This blog — reapbountifully.com. And nine days in, I have real numbers to share. They’re not impressive. But they’re real, and that’s the point.
My Real Numbers After 9 Days (Yes, They’re Small)
Let me be completely transparent. Here’s where this project stands as of day 9:
- Posts published: ~20 posts in 9 days
- Total visitors (Google Analytics): 65 users, 367 pageviews
- Daily visitors: 0–5 on most days
- Countries reaching: Mostly US (my target market), some Korea, Poland, Canada
- Google Search Console: Site indexed, starting to show impressions
- Revenue: $0. Zero. Nothing.
- Time spent per post: About 2 minutes
I have exactly $0 in revenue and sometimes 5 visitors a day. Here’s why I’m still doing it — and why I think this time is actually different.
What This Project Actually Costs Me
One of the biggest mistakes in my previous side projects was not knowing (or caring about) the real cost. This time I tracked everything from day one.
| Item | Monthly Cost | What It Does |
|---|---|---|
| Vultr VPS (Cloud Compute) | $6 | Hosts WordPress, runs 24/7 |
| Claude Pro subscription | $20 | AI writes all content via Claude Code |
| Domain (reapbountifully.com) | ~$1 | Annual cost averaged monthly |
| Total | $27/month | Full blogging operation |
$27 a month. That’s less than a Netflix subscription and a couple of coffees. Compare that to the SaaS side projects I used to dream about — where you need months of development, hosting infrastructure, payment processing, customer support, and marketing budget before you see a single penny.
Why This Time Is Different: 3 Things I Actually Did
This isn’t theory. These are the three specific decisions I made that broke my old pattern of building, abandoning, and repeating.
1. I Chose Monetization First, Then Built Everything Around It
With my old projects, monetization was always “something I’ll figure out later.” With this blog, I decided on AdSense before I wrote a single post. That one decision changed everything:
- Content topics? Chosen for search traffic potential, not what’s technically interesting to me
- SEO? Automated from day one — every post gets a Yoast focus keyphrase and meta description set by AI
- Post frequency? Maximized because more indexed pages = more potential search traffic = more ad revenue
When you know HOW you’ll make money, every other decision becomes obvious. When you don’t, you waste months building features nobody asked for.
2. I Shipped in 30 Minutes, Not 3 Months
My first post went live the same day I set up the VPS. No custom theme tweaking for weeks. No “perfect” design. No building a custom CMS from scratch because “WordPress isn’t developer enough.” (Yes, I’ve done that before.)
Here’s my actual launch timeline:
- Minute 0–10: Spun up a $6 Vultr VPS, installed WordPress
- Minute 10–20: Configured basic theme, installed Yoast SEO
- Minute 20–30: Published first post
- Day 1–3: Built a Telegram bot that lets me write and publish posts from my phone
- Day 4–9: Published 20 posts while riding the subway, waiting in line, during lunch breaks
The blog was live and earning (well, accumulating) SEO value within 30 minutes. Every day of delay is a day Google isn’t indexing your content.
3. I Automated the Boring Parts (So I Can’t Get Bored)
This is the real game-changer, and it’s the reason I’m still publishing on day 9 instead of abandoning the project on day 3 like usual.
My tech stack:
- WordPress — boring, reliable, SEO-friendly. Exactly what you want.
- Claude Code — AI writes full blog posts based on my topic and angle
- Custom Telegram bot — I type a topic on my phone, the bot generates and publishes the post
- /trending command — the bot suggests topics by pulling from Google Trends
- Automated SEO — Yoast focus keyphrases, meta descriptions, all set automatically by AI
The result? A full blog post — written, SEO-optimized, formatted, and published — takes me about 2 minutes. I type a topic into Telegram. The bot does the rest. I review and hit publish.
This matters because the #1 reason my side projects die is boredom with repetitive tasks. I love building systems. I hate doing the same thing manually every day. So I built a system that does the repetitive part, and now publishing is as easy as sending a text message.
The Math That Keeps Me Going
Here’s the simple compound math that makes this project worth the $27/month even with zero revenue today:
If each blog post eventually brings in just 10 organic visitors per month from Google (a very conservative number for a decent post), then:
- 20 posts (now) = 200 visitors/month potential
- 50 posts (in ~2 months) = 500 visitors/month
- 100 posts (in ~4 months) = 1,000+ visitors/month
- 200 posts (by end of year) = 2,000+ visitors/month
At typical AdSense RPMs for tech/developer content ($5–15 per 1,000 pageviews), 2,000 visitors/month means roughly $10–30/month in ad revenue. Not life-changing. But it would cover the $27/month cost, making the entire operation free — and everything after that is profit.
The key insight: content compounds but code doesn’t. My old Telegram bots required constant maintenance to stay relevant. Each blog post, once indexed by Google, brings traffic for months or years with zero additional effort. That’s the difference.
Honest Assessment: What’s Working and What’s Not
I promised real numbers, so here’s my honest scorecard at day 9:
What’s Working
- Consistent publishing: 20 posts in 9 days. I’ve never sustained this pace on any project.
- Google indexing: Search Console shows the site is being crawled and indexed. Impressions are starting to appear.
- Content quality: AI-written posts are genuinely helpful and well-structured. Way better than what I’d write manually in English as a non-native speaker.
- The Telegram bot workflow: Publishing from my phone means I actually do it. Removing friction is everything.
- Cost control: $27/month is sustainable indefinitely. No pressure to monetize immediately.
What’s Not Working Yet
- Traffic: 0–5 daily visitors. SEO takes 3–6 months to really kick in, so this is expected but still humbling.
- Revenue: $0. Haven’t even applied for AdSense yet — need more content and traffic first.
- Some scheduled posts missed: The automation isn’t perfect. A few scheduled posts didn’t go out on time.
- Topic selection: Still figuring out the right balance between what I know (dev tools, AI) and what gets search traffic.
The Compound Effect: Why This Project Survives Where Others Died
Looking at my dead project graveyard, every single one failed for the same reason: the effort-to-reward ratio got worse over time. Building a bot is fun. Maintaining it is not. Adding features is exciting. Fixing bugs for zero users is soul-crushing.
This blog project has the opposite dynamic. The effort per post is nearly zero (2 minutes). But the cumulative reward grows every week as Google indexes more pages. The effort-to-reward ratio gets better over time, not worse.
Plus, there’s something uniquely motivating about watching Google Search Console numbers tick up. Every new impression, every new indexed page — it’s like watching a tiny snowball start to roll downhill. It might take months to become an avalanche, but the direction is clear.
As a Korean developer writing an English blog with AI, I’m essentially removing three major barriers at once:
- Language barrier: Claude writes native-quality English that I could never produce this fast
- Time barrier: 2 minutes per post means I can publish while commuting
- Motivation barrier: The system is so frictionless that “I don’t feel like it today” doesn’t apply — it takes less effort than deciding what to eat for lunch
What I’d Tell My Past Self
If I could go back and talk to the version of me who was building aqt-bs-bot at 2 AM, convinced it would be “the one that takes off,” I’d say this:
Stop building products. Start building assets.
A Telegram bot is a product. It needs users, updates, and maintenance. A blog post is an asset. Once published and indexed, it works for you while you sleep. One hundred blog posts is one hundred tiny employees working 24/7, each one potentially bringing in a few visitors a month, forever.
Will this blog make me rich? Almost certainly not. Will it cover its own costs and teach me more about SEO, content, and online business than any side project I’ve ever built? It already has.
And if you’re a developer with your own graveyard of abandoned repos, maybe the answer isn’t building something more technically impressive. Maybe it’s building something embarrassingly simple — and actually shipping it.
My blog cost $27/month, takes 2 minutes per post, and has made exactly $0 so far. Ask me again in 6 months.
How This Post Was Made
This post was written using Claude Code (AI) based on my real project data and experience. I provided the actual numbers from Google Analytics, Search Console, and my server costs. The AI structured and wrote the post, and I published it via my custom Telegram bot. Total time: about 2 minutes. The irony of writing about automation using automation is not lost on me.
Transparency: This blog uses AI-assisted content creation. All experiences, numbers, and opinions are real — the writing assistance is the AI part. Learn more about this project.