Appearance
Chapter 9 — 30-Day Roadmap
🗓️
"30 days is enough to ship 1 agent with users.Not enough to get rich. Enough to prove you can ship."
How to use this roadmap
- 30 days zero → ship 1 real agent with real users
- 3 tracks by goal: Solo dev | Agency | Operator non-dev
- 2-4 hours/day dedicated
- Daily checkbox
00 Pre-flight (Day 0)
Done before Day 1
- [ ] Claude Code Pro ($20/month) or Cursor Pro
- [ ] Anthropic API key ($5 minimum top-up)
- [ ] GitHub account + 1 empty repo (private OK)
- [ ] VPS or cloud account (Vercel/Cloudflare free OK)
- [ ] Twitter/X account for build-in-public
- [ ] Tracking sheet (Notion / Apple Notes)
- [ ] Calendar block 2-4h/day same time
🛤️ Pick 1 of 3 tracks
| Track | Day 30 Goal | Effort |
|---|---|---|
| A. Solo dev | Ship 1 agent product with 5+ paying customers | 3-4h/day |
| B. Agency | Close 1 deal $5-10K for SME | 2-3h/day + sales meetings |
| C. Operator | Automate 5 internal workflows, save 10h/week | 2h/day |
TRACK A — Solo dev (agent product with users)
Week 1 — Setup + idea (Day 1-7)
| Day | Task |
|---|---|
| 1 | Brainstorm 10 agent product ideas (Chapter 5 + 6 inspiration) |
| 2 | Pick 1: niche + clear pain + identifiable audience |
| 3 | Spec with Claude: feature list, user flow, tech stack |
| 4 | Landing page (Carrd / Framer) + waitlist form |
| 5 | Post Twitter / Reddit / FB group → 50+ signups |
| 6 | Customer dev: call 5 users, ask 5 questions |
| 7 | Refine spec based on interviews |
Week 2 — Build core (Day 8-14)
| Day | Task |
|---|---|
| 8 | Setup repo: Next.js + Supabase + Stripe + Clerk + Claude API |
| 9 | Auth flow + landing → dashboard |
| 10 | Core agent loop (input → LLM + tool → output) |
| 11 | Save history to DB + display dashboard |
| 12 | Payment integration (Stripe checkout + webhook) |
| 13 | Polish: empty state, error, loading |
| 14 | Internal dogfood — 50 generations own use |
Week 3 — Launch (Day 15-21)
| Day | Task |
|---|---|
| 15 | Pricing page (3 tiers) + Privacy + Terms |
| 16 | Email waitlist: "we're live!" |
| 17 | ProductHunt + Hacker News launch |
| 18 | Reddit + Twitter thread + FB group |
| 19 | Reply to every comment, fix critical bugs |
| 20 | Customer support setup (email + Intercom) |
| 21 | Week recap: signups, conversion, MRR |
Week 4 — Iterate + scale (Day 22-30)
| Day | Task |
|---|---|
| 22-24 | Ship 1 feature based on feedback (2 days/feature) |
| 25 | Pricing A/B test |
| 26 | Cold outreach 50 (Twitter DM + email) |
| 27 | Add referral system (free credit for invite) |
| 28 | Cohort analysis: week-1 retention |
| 29 | Eval suite: 30 test cases for agent |
| 30 | Recap: MRR, paying customers, learnings, next 30 |
Day 30 KPI
- Agent product live + payment works
- 5+ paying customers
- $50-500 MRR
- 100+ free users
- Eval suite + cost monitoring active
TRACK B — Agency (SME deal)
Week 1 — Skills + portfolio (Day 1-7)
| Day | Task |
|---|---|
| 1 | Setup n8n cloud + Smax.ai trial + Claude API |
| 2 | Build hello-world workflow (Slack mention + Claude reply) |
| 3 | Build Smax.ai sample bot (Messenger sales) |
| 4 | Build voice agent demo (Vapi free credit) |
| 5 | Compile 3 demos into portfolio site |
| 6 | Write 1 case study post (LinkedIn + blog) |
| 7 | Identify 20 target SMEs (industry: F&B, fashion, retail) |
Week 2 — Outreach + meetings (Day 8-14)
| Day | Task |
|---|---|
| 8 | Cold message 20 SMEs (Facebook DM, email, LinkedIn) |
| 9 | Follow-up + qualify first 5 replies |
| 10-11 | Discovery call 3 SMEs — understand pain |
| 12 | Draft proposals for 2 strongest leads |
| 13 | Send proposal + follow-up call |
| 14 | Recap: meeting count, pipeline value |
Week 3 — Close + deliver MVP (Day 15-21)
| Day | Task |
|---|---|
| 15 | Negotiate + close 1 deal ($5-10K) |
| 16 | Kick-off meeting with client, requirements |
| 17-19 | Build MVP (Smax.ai or n8n based) |
| 20 | Client review + iterate |
| 21 | Train client team (1h session) |
Week 4 — Handover + recurring (Day 22-30)
| Day | Task |
|---|---|
| 22 | Final delivery + documentation |
| 23 | Pitch recurring support contract ($300-1K/month) |
| 24 | Get client testimonial / case study consent |
| 25 | Publish case study (blog + LinkedIn + Twitter) |
| 26-28 | Generate more leads: cold + warm referrals |
| 29 | 2 more discovery calls |
| 30 | Recap: total revenue, recurring, next-month pipeline |
Day 30 KPI
- 1 deal closed $5-10K
- 1 recurring contract active
- Portfolio + 1 case study published
- Pipeline 5+ qualified leads next month
TRACK C — Operator non-dev (workflow automation)
Week 1 — Audit + setup (Day 1-7)
| Day | Task |
|---|---|
| 1 | List 10 tedious tasks you do daily |
| 2 | Rank by time spent + frustration |
| 3 | Pick 5 tasks that can be automated |
| 4 | Setup n8n cloud Starter ($20) or Lindy free tier |
| 5 | Setup ChatGPT Plus ($20) or Claude Pro |
| 6 | First automation: 1 simple task |
| 7 | Test 1 week, measure time saved |
Week 2 — Build 5 workflows (Day 8-14)
| Day | Task |
|---|---|
| 8-9 | Workflow #2: Email triage (Claude classify + label) |
| 10-11 | Workflow #3: Cross-app data sync (Google Sheet → Notion) |
| 12-13 | Workflow #4: Calendar assistant (book + reminder) |
| 14 | Workflow #5: Daily digest (news + email → summary) |
Week 3 — Refine + measure (Day 15-21)
| Day | Task |
|---|---|
| 15-17 | Daily run all 5 workflows, fix edge cases |
| 18 | Measure: time saved/week vs baseline |
| 19 | Add error handling + Slack alerts |
| 20 | Document each workflow (mini-runbook) |
| 21 | Week 3 recap |
Week 4 — Expand + share (Day 22-30)
| Day | Task |
|---|---|
| 22-24 | Pitch boss / team: try workflow tools for them |
| 25-26 | Build 2 workflows for colleagues (free, build goodwill) |
| 27 | Write internal doc: "5 workflows I automated" |
| 28 | Share LinkedIn + company blog |
| 29 | Quantify ROI: time × hourly rate |
| 30 | Recap: time saved/week, colleagues onboarded |
Day 30 KPI
- 5 workflows active
- 10+ hours/week saved
- 2 colleagues onboarded
- Doc + share
Common across 3 tracks
Regardless of track
1. Daily build-in-public
- 1 tweet/day sharing progress
2. Weekly retro (Sunday)
- KPI check
- Top 3 learnings
- Next week priorities
3. Tool budget cap
- Start < $100/month
- After $500+ MRR: increase
4. Energy management
- 2-4h focus, no multitask
- 1 day off/week
Common pitfalls
🚨 8 stumbles
1. Switch track mid-way → reset Day 1. Stick 30 days. 2. Perfectionism Week 1 → don't wait for "pretty" MVP. Ship Week 3 even ugly. 3. Over-tool → 10 tools, use 2. Stick 3-5 core. 4. Skip build-in-public → start Day 1, not Day 30. 5. Compare with top creators → they took 5 years, you're on Day 7. Apple-to-apple. 6. Forget specific KPI numbers → "improving" = not measurable. Need number target. 7. Burnout week 3 → trough of disillusionment. Recharge, don't quit. 8. No distribution plan → ship but no one knows. Build distribution Day 1.
Global tips
Time zone
- Emerging market UTC+5 to +9 — generate while US sleeps
- Post during US morning (8-9am EST) = peak engagement
Communities
- F8 community (Vietnam — Sơn Đặng)
- IndieHackers — global
- Vietnam Tech Twitter (#vntech)
- Built in Saigon community
When to quit day job?
| MRR | Action |
|---|---|
| < $500 | Side project |
| $500-2K | Part-time side |
| $2K+ | Can quit (livable in many markets) |
| $5K+ | Definitely quit / scale team |
Tracking template
Day | Date | Task done | Energy (1-10) | KPI today | Tomorrow
─────────────────────────────────────────────────────────────
1 | 1/6 | Spec idea | 8/10 | 0 | Landing
2 | 2/6 | Landing | 7/10 | 12 signup | Customer dev
...After Day 30
- Chapter 2 — Claude Code Deep — refine workflow
- Chapter 5 — Workflow Agent — scale
- Chapter 6 — MCP — build moat
- Chapter 8 — Safety — production-ready
Final word
"30 days isn't enough to get rich.30 days is enough to prove you can ship.90% of people don't finish Week 1.You just need to be in the 10% that does.Day 1 starts now. Go ✊"