100 Posts on dev.to: 10 Things I Learned Writing About Indie Dev

I just hit 100 posts on dev.to — Flutter, Supabase, AI APIs, and indie dev workflows. Here's what surprised me along the way.

The 100 Posts Breakdown

By category:
  Flutter:          22 posts (state mgmt, animations, testing, Web perf)
  Supabase:         18 posts (Auth, RLS, Realtime, Edge Functions)
  AI/LLM:           20 posts (Claude API, prompt design, agents)
  Indie dev:        25 posts (KPIs, SEO, costs, marketing)
  GHA/automation:   15 posts (CI/CD, scheduling, security)

By language:
  English (EN): 50 posts (dev.to international)
  Japanese (JA): 50 posts (drafts toward Qiita/note)

10 Things Writing 100 Posts Taught Me

1. A Published Post Beats a Perfect Draft

My first 10 posts sat in drafts for weeks — "not quite ready yet." Looking back from post #100, quality grew linearly with volume. Done is better than perfect.

2. Write While Implementing — Not After

Bad pattern:
  implement → it works → "I'll write about it later" → 3 weeks pass → can't remember

Good pattern:
  write as you implement →
  every sticking point becomes the "problem" section of the post

3. Code Blocks Drive Reads

Analytics showed posts with code blocks get 2× the reads. A Flutter state management example I wrote "just for reference" turned into my most-visited post.

4. Specific Numbers in Titles Win on CTR

❌ "Flutter state management comparison"
✅ "Flutter State Management Compared: Provider vs Riverpod vs Bloc"

❌ "Cut indie dev costs"
✅ "Indie Dev Cost Management: Designing a $0→$100/Month Stack"

5. English Posts Reach a Global Audience

My English posts got 5× the impressions of Japanese ones. Technical writing translates well even with imperfect English — the code communicates.

6. Failure Stories Get Shared Most

Posts about mistakes ("3 products that failed", "10 lessons from failure") consistently outperformed success stories in shares. Failure is relatable.

7. Automation Keeps the Habit Alive

# Weekly reminder if unpublished drafts pile up
on:
  schedule:
    - cron: '0 9 * * MON'
jobs:
  blog-reminder:
    steps:
      - run: |
          COUNT=$(find docs/blog-drafts -name "*-en.md" \
            | xargs grep -l "^published: false" | wc -l)
          echo "Unpublished drafts: $COUNT"

Building a system that reminds me when drafts stack up cut "too busy to post" gaps significantly.

8. Tags Matter — dev.to Silently Caps at 4

dev.to only applies the first 4 tags from your frontmatter — extra tags are dropped without warning. Always put specific tags (flutter, supabase) before generic ones (webdev, programming).

9. Series Format Increases Completion Rate

Writing "Phase X" series posts is far easier to sustain than standalone posts. Zero decision cost for "what do I write next." The series structure carries you forward.

10. Writing Forces Real Understanding

Every time I tried to write about something I "understood," I discovered gaps. Supabase RLS row policies and Claude API Prompt Caching both only became fully clear to me when I had to explain them in writing.

Toward the Next 100

Goal: 200 posts on dev.to (within 1 year)
New topics:
  - AI agent production case studies
  - Flutter + Supabase real-world ops lessons
  - Indie SaaS monetization in the open

More automation:
  - GHA weekly draft generation
  - Auto-measure SEO impact post-publish

100 is a checkpoint. The next 100 will focus on tightening the "write → get read → improve" cycle.

---

All 100 posts: dev.to/kanta13jp1