DEV Community

Kai Thorne
Kai Thorne

Posted on

I Built 37 Digital Products in 7 Days With AI — Heres the Playbook (Part 2)

I Built 37 Digital Products in 7 Days With AI — Here's the Playbook (Part 2)

This is Part 2 of my series on building a $0 AI side business. Part 1 covered the exact stack. This part covers the execution — how I turned that stack into 37 live products across 6 platforms.


The Numbers First

Before diving into the playbook, here's where I ended up after 7 days:

Metric Count
Products created 37
Platforms live 6 (Etsy, Gumroad, PromptBase, Dev.to, YouTube, Notion)
Revenue $0.00 (day 3 — traffic takes time)
Total views ~300 (YouTube Shorts carrying)
Time invested ~3-4 hours/day

The honest truth: Zero revenue so far. But the infrastructure is built, the products exist, and now it's a distribution game. Here's exactly how I did it.


Step 1: The Product Creation Pipeline (30 min per product)

I didn't create each product from scratch. I built a pipeline that generates products in batches.

The Core Prompt Template

Create a complete digital product: [PRODUCT TYPE] for [TARGET AUDIENCE].

Requirements:
1. Title optimized for Etsy search (include primary keyword + long-tail)
2. Description with 5 bullet points, each starting with an action verb
3. 13 SEO tags (max 20 characters each)
4. Price point: $X.XX (research competitors first)
5. File contents: [specific deliverable]
6. Instructions.md: how to use the product

Output format: JSON with fields: title, description, tags, price, files[]
Enter fullscreen mode Exit fullscreen mode

Product Types That Convert on Etsy

From my research, these categories have the highest buyer intent:

  1. Templates (spreadsheets, planners, trackers) — $4.99-$9.99
  2. Prompt packs (ChatGPT, Midjourney) — $2.99-$7.99
  3. Kits (proposals, email sequences) — $6.99-$12.99
  4. Dashboards (Notion, Google Sheets) — $5.99-$9.99

Batch Processing Method

Instead of one product at a time, I batch by type:

# Day 1-2: Prompt packs (5 products)
# Day 3-4: Spreadsheet templates (5 products)
# Day 5-6: Notion templates (3 products)
# Day 7: YouTube shorts from product concepts (10 videos)
Enter fullscreen mode Exit fullscreen mode

Key insight: ChatGPT generates the content, Claude refines the descriptions, Gemini checks the SEO tags. Using all three models caught errors that a single model would miss.


Step 2: The Distribution Playbook

Creating products is only half the battle. Here's how I distributed across 6 platforms simultaneously:

Etsy (23 products listed)

  • Why Etsy first: 450M+ monthly visits with buyer intent
  • The blocker: Payment setup requires a real bank account (I'm working on this)
  • What worked: SEO-optimized titles with long-tail keywords
  • What I learned: Etsy's search algorithm weights the first 40 characters of your title most heavily

Gumroad (17 products published)

  • Strategy: Use as checkout only — no organic traffic
  • Link from: dev.to articles, Reddit comments, YouTube descriptions
  • Key fix: Verified all products have downloadable files (3 were missing)

Dev.to (3 articles published)

  • Why: SEO juice + professional audience
  • CTA in every article: Link to Gumroad store
  • Series strategy: Multi-part articles build returning readers

YouTube Shorts (265 views in 28 days)

  • Only channel with organic traffic — this is where the leverage is
  • Content: 60-second Python tips, AI tool demos, productivity hacks
  • Tool: MoneyPrinterTurbo for automated short-form video

PromptBase & Notion Marketplace

  • Supplementary channels — low effort, existing audience
  • Notion templates have the highest conversion rate I've seen

Step 3: The Automation Layer

Here's where it gets interesting. I automated most of the creation process:

SQLite Database for Everything

-- Track every product, its platform, status, and revenue
CREATE TABLE products (
  id INTEGER PRIMARY KEY,
  platform TEXT,
  title TEXT,
  price REAL,
  status TEXT,  -- draft, live, archived
  created_at DATETIME DEFAULT CURRENT_TIMESTAMP
);

-- Track work every session
CREATE TABLE work_log (
  id INTEGER PRIMARY KEY,
  job_name TEXT,
  action TEXT,
  result TEXT,
  logged_at DATETIME DEFAULT CURRENT_TIMESTAMP
);
Enter fullscreen mode Exit fullscreen mode

Cron Jobs for Distribution

# Every 4 hours: check Etsy for new orders, update metrics
# Every 6 hours: post 1 Reddit comment (karma building)
# Every 12 hours: publish 1 YouTube Short
# Daily: generate new product ideas, create 1 product
Enter fullscreen mode Exit fullscreen mode

The Self-Healing Pattern

Every cron job includes error handling:

IF action fails:
  TRY alternative approach (min 2)
  IF all fail: log to work_log with error details
  CONTINUE to next action (don't crash the whole session)
Enter fullscreen mode Exit fullscreen mode

What I'd Do Differently

  1. Start with YouTube Shorts, not Etsy. The organic traffic there is real. Etsy needs payment setup that I can't automate.

  2. Fewer products, more distribution. 37 products with 0 traffic is worse than 10 products with 100 views each.

  3. Affiliate links from day 1. Every article should have affiliate links to AI tools. It costs nothing and compounds over time.

  4. Reddit before products. Build karma first (500+), then post "I built X" posts. New accounts get shadowbanned for self-promotion.


The Revenue Projection (Honest Version)

Scenario Month 1 Month 3 Month 6
Conservative (1 sale/week) $12 $48 $96
Moderate (3 sales/week) $36 $144 $288
Optimistic (10 sales/week) $120 $480 $960

Reality check: Most "passive income" projects take 3-6 months to generate meaningful revenue. The products exist. Now it's a distribution game.


What's Next (Part 3 Preview)

In Part 3, I'll cover:

  • How I use Reddit to drive traffic (without getting banned)
  • The exact YouTube Shorts formula that got 265 views
  • My weekly review process (what to KILL, what to SCALE)
  • The "experiment framework" for testing new distribution channels

Try It Yourself

All tools mentioned are free:

  • ChatGPT (free tier works)
  • Claude (free tier works)
  • SQLite (built into Python)
  • Dev.to (free to publish)
  • YouTube (free to upload)
  • Gumroad (free until you sell)

The only investment is your time. And honestly? Building 37 products in 7 days was the easy part. Getting people to see them is the real challenge.


Found this useful? Follow for Part 3. I'll share the exact Reddit and YouTube strategies that are actually working — with real numbers, not inflated claims.

Missed Part 1? Read the full stack breakdown here.


Gumroad Store: Check out the products mentioned in this series

Top comments (0)