Start with an idea you can test
Maybe you want to launch something on the side but don’t know where to start. Or you’ve got a list of ideas and keep switching between them.
Either way, you don’t want to spend 3 months building something nobody wants.
ChatGPT can help you narrow the list and work out what to test first. Treat its suggestions as a starting point, though. A convincing answer isn’t proof that customers will pay.
Step 1: Set clear filters
“Give me 10 business ideas” leaves ChatGPT guessing about your budget and what you’re willing to build.
Give it some limits:
“Give me 5 business ideas that:
require < $500 to start
solve a real problem
use automation or AI
are not dropshipping
can be tested in 30 days
have clear monetization (subscription, course, tool, etc.)”
Swap in your own constraints. If you only want to sell to businesses, say so. If you already know how to reach creators, ask for ideas that serve them. The suggestions should fit your situation.
Step 2: Compare the ideas
Next, ask ChatGPT to put the ideas side by side. Include your actual strengths in the prompt; the example below assumes you’re a good writer and marketer.
“Now put these 5 ideas into a table and score them 1–10 for:
Market Size
Monetization Potential
AI Leverage
Ease of Execution
Personal Fit (for someone who’s a good writer and marketer)”
Example comparison
These scores are illustrative, not market research. Use the table to organize your thinking, not to decide which business will succeed.
Idea | Market Size | Monetization | AI Leverage | Ease of Exec | Personal Fit | Total |
|---|---|---|---|---|---|---|
AI-powered podcast clipping tool | 9 | 7 | 8 | 6 | 6 | 36 |
Local biz review summary | 7 | 8 | 9 | 8 | 7 | 39 |
Shopify product copy with AI | 9 | 9 | 8 | 9 | 9 | 44 |
Personal news briefings | 6 | 5 | 10 | 7 | 9 | 37 |
AI property listing generator | 8 | 9 | 7 | 6 | 6 | 36 |
In this example, Shopify product copy comes out on top. Before you commit to it, ask:
“What are the biggest risks or weaknesses with this top idea?”
Use the answer to identify what you need to check with potential customers. The highest score doesn’t settle the decision.
Step 3: Plan a small first version
Once you’ve picked an idea to test, ask for a minimum viable product (MVP) plan. Keep the budget and deadline in the prompt:
“Create a lean MVP plan for this Shopify product description generator. Include tools, steps, and timelines over 3 weeks. Budget = $500.”
ChatGPT might suggest tools like these. You won’t need all of them, and you’ll need to check current pricing against your budget:
Carrd / Typedream → quick site
Tally / Typeform → collect leads
Make / Zapier / Notion / Airtable → backend
ChatGPT / Claude / Gemini → content generation
Stripe → payments
Then prompt:
“Now write the first 5 tweets + 5 LinkedIn posts to promote it.”
You now have a draft plan and some promotional posts to edit. The next useful step is to put the offer in front of potential customers and see how they respond.
Bonus prompt: Look at the competition
“Now act like a competitor. What product would you build to beat me in 6 months?”
This can surface weaknesses you haven’t considered. Treat the answer as a list of possibilities to investigate, not a prediction.
What to do next
Start with your constraints, compare a few ideas, and question the one you like best. Then plan the smallest version you can test within your budget.
Don’t spend weeks polishing the scores. Use them to choose a test, then let customer feedback shape what you build.