AM Choosing a niche is the decision that shapes everything after it: the keywords you can target, the products you can sell, how quickly you can publish and how much each visitor is worth. A good niche makes an average site profitable. A bad niche makes a well-built site struggle for years.
This guide gives you a practical process for how to find a niche, whether you plan a content site (blog, affiliate or ad-supported) or an online store. You will generate ideas, score them with a simple table, validate the best one in about a week and check it against the red flags that sink most new projects.
What makes a niche worth building
A niche is a topic narrow enough that you can cover it better than general sites, and broad enough that it has hundreds of real questions and a way to earn money. "Fitness" is not a niche. "Kettlebell training at home for beginners" might be. "Pets" is too wide. "Care and gear for backyard chickens" is workable.
For content sites, the niche needs search demand across many specific questions and a monetization path such as affiliate programs, display ads or a digital product. For stores, it needs products with healthy margins, reliable suppliers and customers who search for what you sell. Many of the best projects are both: a store with a content section, or a content site that later adds products.
Where to find niche ideas
Aim for 20 to 30 raw ideas before judging any of them. Quantity first, filtering later. Good sources:
- Your own life. Hobbies, jobs, past purchases you researched for weeks, problems you solved. Knowledge you already have is a real advantage, because it shows in the content.
- Marketplace categories. Browse the category trees and bestseller lists on Amazon, Etsy and large niche retailers. Each subcategory is a possible niche. Look for ones with many products and active reviews.
- Communities. Subreddits, Facebook groups, Discord servers and forums show what people care about enough to discuss daily. Busy communities with repeated beginner questions are a strong signal.
- Magazines and YouTube. If a topic supports a print magazine or several channels with steady views, it has an audience that spends money.
- Google Trends. Check a topic's five-year line. You want steady or rising interest, not a single spike.
- Adjacent niches. Take a broad market and look for groups it serves badly: left-handed tools, small-space gardening, gear for tall people, gluten-free baking at high altitude.
- Your own purchase and search history. Scroll back through a year of orders and searches. Anything you spent hours researching before buying is a topic where other buyers need help too.
- Problems at work. Trades and professions have tools, software and processes that outsiders never think about. If you know a field from the inside, list the questions new people in it always ask.
Write every idea down in a spreadsheet with one row each, even the ones that seem weak. Some of the best niches start as a half-formed note that only makes sense once you compare it with the others.
For store ideas, the dropshipping niches guide covers product categories and supplier questions in more depth.
How to narrow a broad idea
Most first ideas are too wide. "Gardening", "cooking" and "home fitness" each have thousands of pages from large publishers. You do not need to abandon a broad idea, you need to cut it down to a corner you can own. Three ways to do that:
- By audience. Gardening for renters, cooking for one person, fitness for people over 60. A specific reader gives every page a clear angle.
- By situation or constraint. Balcony gardening, cooking without an oven, home workouts without equipment. Constraints create questions that general sites answer poorly.
- By method or product type. Hydroponic gardening, cast iron cooking, kettlebell training. A method has its own gear, techniques and problems, which means depth.
A useful formula is audience plus problem plus method: "container vegetable gardening for apartment balconies". It sounds small, but write out the questions such a person has (which vegetables fit, how deep the pots must be, how much sun is enough, how to water while traveling, which fertilizer to use) and the list grows quickly. You can always widen the site later once it ranks. Starting narrow and expanding is far easier than starting wide and trying to stand out.
Score each idea on six criteria
Once you have a list, score each idea from 1 (poor) to 5 (strong) on the six criteria below. Be honest. The point is to compare ideas against each other, not to prove your favorite is perfect.
| Criterion | What you are checking | Scores 1 | Scores 5 |
|---|---|---|---|
| Demand | Do people search this topic across many specific questions? | Few autocomplete suggestions, tiny communities | Hundreds of real questions, active communities, steady trend |
| Competition | Can a new site rank? | Page one is all major brands and large publishers | Forums, thin pages and small sites rank for many queries |
| Monetization | How does each visitor turn into money? | No products, no programs, low-value audience | Several affiliate programs, products with good margins, buyers with budget |
| Your knowledge | Can you add real experience? | You would be summarizing other sites | You have used the products or done the work for years |
| Content depth | Is there enough to write about? | Under 30 useful pages before you repeat yourself | 100 or more distinct topics with room to grow |
| YMYL risk | Could bad advice harm health, money or safety? | Medical, legal, financial or safety advice | Hobbies, home, crafts, most product categories |
YMYL stands for "Your Money or Your Life", a term from Google's Search Quality Rater Guidelines for topics that could significantly affect a person's health, financial stability, safety or welfare. Google holds these pages to a higher standard of trust and expertise. A new site without credentialed authors will find YMYL topics very hard, so a high score here means low risk.
Example scoring
| Idea | Demand | Competition | Monetization | Knowledge | Depth | YMYL | Total |
|---|---|---|---|---|---|---|---|
| Backyard chickens | 4 | 3 | 4 | 4 | 5 | 4 | 24 |
| Home espresso | 5 | 2 | 5 | 3 | 4 | 5 | 24 |
| Diabetes meal planning | 5 | 2 | 3 | 2 | 5 | 1 | 18 |
| Vintage typewriter repair | 2 | 4 | 2 | 5 | 3 | 5 | 21 |
If you care more about one factor, weight it. For a first site, doubling the competition score makes sense, since ranking at all is the first hurdle. Treat any single score of 1 as a warning, even if the total looks good. In the example, the diabetes idea scores well on demand but its YMYL and knowledge scores make it a poor first project.
Validate your top two ideas
Scoring uses judgment. Validation uses evidence. Take your top two ideas through the steps below, which take about a week per idea, before committing to one.
- Mine real searches. Pick 10 to 15 seed terms and run them through Google autocomplete from A to Z, plus question words like how, why and can. If you end up with fewer than 200 distinct, relevant keywords, the niche may be too thin. The method is laid out in niche keyword research.
- Read the results for 20 keywords. Choose a mix of questions and buyer searches. For each, note who ranks. If you find forum threads, outdated posts or small sites in the top five for at least a third of them, there is room.
- Find the money. List at least three ways to earn: specific affiliate programs and their commission terms, products you could sell and their likely margins, or digital products the audience already buys. If you cannot name three, monetization is a guess.
- Check the audience's spending. Are people in the communities buying gear, courses or services? Posts like "what should I buy" and "is this worth it" are good signs.
- Outline 50 page titles. If you struggle past 30 without repeating yourself, content depth is lower than you scored it.
- Look at small competitors. Find two or three small sites ranking in the niche. If sites similar to what you would build are getting traffic, the model works here.
The Niche Research and Keyword Miner in AM Jarvis speeds up steps 1 and 2: it mines autocomplete across 18 countries, groups the results into a topical map and shows an "untapped" score based on the current results, so you can compare two niches side by side from real search data.
Extra checks for store niches
A store has costs a content site does not, so run these checks too:
- Margin after all costs. Subtract product cost, shipping, payment fees, platform fees and an allowance for returns and ads. The Shopify fee calculator helps you see the platform and payment side.
- Size and weight. Large or heavy items eat margin in shipping and make returns expensive.
- Fragility and returns. Clothing and shoes have sizing returns. Glass and electronics have damage claims. Price this in.
- Supplier reliability. Order samples. Check shipping times to your main market.
- Repeat purchase. Consumables and accessories bring customers back. One-time purchases need constant new traffic.
- Ad and Merchant Center restrictions. Supplements, weapons-related items and some health products face advertising limits or extra review on Google and social platforms. Check the policies before you choose.
Red flags that should stop you
Warning: One red flag is enough reason to pause. Two or more usually mean you should move to the next idea on your list.
- A single spike on Google Trends. Fads collapse. If the interest is one peak from a viral moment, the niche may be empty in a year.
- Page one is only big brands for almost every query. If you checked 20 keywords and saw no forums, small sites or weak pages at all, a new site will wait a very long time.
- You cannot name how it makes money. "I will add ads later" is not a plan. Many hobby niches have large audiences that rarely buy.
- Heavy YMYL without real expertise. Health, finance and legal topics need credentials and careful sourcing. Without them, you are fighting the ranking systems and the risk of misleading people.
- Answers fit in one sentence. If most questions are fully answered in the search results themselves, visitors have little reason to click through.
- You do not care about it. You will write or edit at least 100 pages. Interest keeps you going when traffic is slow for the first months.
- Dependence on one program or supplier. A niche that only works through one affiliate program or one supplier can break overnight when their terms change.
Pick one and commit
After scoring and validation, you will usually have one idea that is clearly ahead. Pick it and commit for at least six months. New sites generally take months to gain steady search traffic, and switching niches early resets that clock. Most failed niche sites did not fail because of the topic. They stopped publishing before the site had a chance.
A simple first-month plan keeps momentum: finish the topical map in week one, publish your first pillar page and three to five cluster pages by the end of week two, set up Google Search Console so you can see impressions from the start, and keep a steady weekly publishing target after that. Write down your validation notes too. If the niche underperforms after six months, those notes show whether the problem was the topic, the competition or the execution.
Your next step is structure. Turn the keywords you mined during validation into a topical map and a page plan using the guide on keyword research for niche sites. If you are building a store, the store name generator can help you find a brand name that fits the niche without boxing you into a single product.