- Generate a list of 20 niche ideas from your interests and AI
- Validate demand, competition and monetization for your top 3
- Choose one niche with a clear audience and problem
Lesson video coming soon. The full written lesson is below.
Your niche decides almost everything that follows: your domain, your content, your products and your traffic channels. In this lesson you will brainstorm widely, then validate ruthlessly, so you only build on a niche with real demand, room to compete and a realistic path to revenue.
What makes a good niche
- A specific audience. “Home cooks” is broad. “Busy parents batch-cooking on Sundays” is a niche.
- Real problems or desires. People search, ask and buy around it.
- Room to be useful. Existing results or products are thin, dated or generic.
- Money signals. Products, affiliate programs, paid tools or buyers already exist.
- Staying power. You can create for it for a year without burning out.
Step by step: from 20 ideas to 1 niche
- Brain dump 10 seeds. List hobbies, job skills, past problems you solved, purchases you researched heavily and questions people ask you.
- Expand to 20 with AI. Use the prompt below to turn seeds into specific audience-plus-problem niches that fit your asset model.
- Cut to your top 3. Remove anything you would not enjoy, anything that needs credentials you lack (medical, legal, financial advice) and anything built around someone else’s trademark.
- Check demand. Use the keyword idea box, Google autocomplete, “People also ask” boxes and Google Trends. For stores, browse marketplace search suggestions. For domains, look at recent public sales in the category. For tools, look for people describing the problem and existing workarounds.
- Check competition. Search 5 core queries. Are results from huge brands only, or do smaller sites and forums show up? Are products all identical? Do existing tools have frustrated reviews?
- Check money signals. Look for affiliate programs, sponsors, price points, paid templates or tools. No money signals is a warning, not always a deal-breaker, but it means you will need your own products.
- Read the audience. Spend 20 minutes on forums, Reddit, YouTube comments and product reviews. Copy exact phrases into your journal.
- Score and choose. Rate each of your top 3 in the Niche Scorer below and pick the winner.
I am building a [niche site / POD store / domain portfolio / micro-SaaS]. My seeds: [list 10 interests, skills or problems]. Generate 20 specific niche ideas. For each give: - Audience (who exactly) - Core problem or desire - 3 example search queries they might type - Likely ways to earn (affiliate, ads, products, subscriptions, domain buyers) - One risk (seasonality, heavy competition, YMYL topic, trademark issues) Avoid niches built on brand names or trademarks. Do not estimate income.
Score your top three
Niche Scorer
Score any niche on 10 factors and compare ideas side by side.
Loading toolβ¦
Prefer the full-page version? Open the Niche Scorer. For a longer walkthrough, read how to find a profitable niche and the find your niche guide.
Validation notes by model
Niche site
You want a cluster of 30+ specific questions you can answer better than current results, including some that AI answers cannot fully satisfy (comparisons, hands-on testing, local or personal experience).
Store / POD
You want an identity-driven audience (profession, hobby, pet, place) that buys expressive products. Never use brand names, characters or logos in designs.
Domains
You want a category where end users (businesses, startups) actively buy names. Study the space; avoid anything close to an existing trademark.
Micro-SaaS
You want a repeated, painful task that people currently solve with spreadsheets or clumsy workarounds, ideally for people who already pay for software.
Red flags that should make you pause
- Every result is a giant brand. If page one is only major publishers and marketplaces for every query, go narrower.
- The whole niche depends on one merchant. One affiliate program or one platform means one policy change can end the model.
- You feel nothing about it. Boredom shows in your content and makes month three very hard.
- It is a fad with no history. Check Google Trends over five years. Spikes are fine as a bonus, not as a foundation.
Generate your 20 ideas, validate your top 3 using steps 4 to 7, score them in the Niche Scorer and write your niche as one sentence: “I help [audience] with [problem] through a [asset model].” Save 10 exact audience phrases in your journal. Next up: Set Up Your AI Builder Stack.