You don't have to guess, and you definitely shouldn't build anything yet. Talk to people first, then pre-sell before you create a single lesson.
1. Pick the person, not the topic. The topic usually falls out of this. Ask who needs your help the most, not who could use it. You want someone at a tipping point with real pain and urgency, not a beginner who's "just curious". Think of someone who needs the language for a specific reason, like relocating for work.
2. Have 10-20 research conversations (aim for about 50 eventually). Start with people you know, past students, and referrals, then move into groups and comments sections. Keep it low pressure: "I'm researching, 30 minutes, no sales." Ask things like:
- What's your biggest frustration with this right now?
- What have you already tried or paid for that didn't work?
- What does success look like for you?
- What would you type into Google or YouTube to solve it?
Use their exact words. Those phrases tell you which of your five topics matters.
3. Sketch a one-page outline. 4-8 modules, 3-5 lessons each. No platform, no website.
4. Pre-sell it to the people you interviewed. Offer it as a founding cohort with a start date and a price based on the value of the outcome. If people pay, you've validated it. If nobody does, you've lost a few weeks, not three months.
5. Deliver it live over 6-8 weeks on Zoom with Google Docs, and record everything. You get paid while you build.
A tip on what makes a topic profitable: "how to speak Italian" is Googleable and cheap, while Italian for people relocating to Italy for work is a prescription people pay real money for.