How to Find a Niche for Your Website: The 5-Step Demand Test
Most first websites don’t fail because of a bad theme or the wrong host. They fail at the niche. People pick a topic nobody searches for, or one where every result on page one belongs to a giant, and then wonder why nothing ranks after six months. So before you buy a domain or touch WordPress, here’s how to find a niche using the same 5-step test I run before I launch any site.
A good website niche passes five checks: you can write 50+ posts on it, people search for it, the top results are beatable, there are affiliate programs or products to earn from, and demand is stable or growing. Score every idea against those five before you buy a domain.
- A niche is three things together: a topic, a specific audience and a way to earn. Miss one and you have a hobby, not a website business.
- Start from what you already know. Your job, studies, hobbies and problems you’ve solved are the best niche ideas you have.
- Check demand with keyword volume and Google Trends, then check competition with Keyword Difficulty and the Domain Rating of the top 10 results.
- Forums and Reddit threads ranking on page one are the best sign a new site can break in.
- No affiliate programs and low CPCs usually mean no money, however much traffic you get.
- If you can’t list 50 post ideas in an hour, the niche is too narrow.
What a niche actually is (topic + audience + way to earn)
A website niche is a specific topic, written for a specific group of people, with a clear way to make money from them. “Coffee” is a topic. “Home espresso for beginners who want café coffee without café prices, earning from machine and grinder affiliate programs” is a niche.
All three parts matter. A topic with a big audience but no way to earn gives you traffic you can’t monetise, which is a nice graph and not much else. A way to earn with no audience searching for it gives you nothing to rank for. You want all three before you buy a domain.
If you’re new to the “way to earn” part, read what affiliate marketing is and how to get started in South Africa first. Most of the sites in this course earn from affiliate links, so it helps to know how commissions work before you judge a niche on them.
Step 1: List what you know
The best niche ideas come from your own life, because you’ll write better content, spot bad advice faster and stick with it longer. Grab a notebook and spend 20 minutes listing:
- Your work: the tools, software and problems you deal with every week.
- Your studies: anything you’ve been trained in, even if you don’t use it now.
- Your hobbies and spending: what you buy, research and compare before you buy it.
- Problems you’ve solved: paying off debt, moving cities, learning a language, starting a side hustle.
- Questions friends ask you: if people come to you for advice on something, strangers on Google want it too.
Aim for 20 to 30 raw ideas. Don’t judge them yet. Steps 2 to 5 do the judging for you. If you get stuck, the Claude prompt further down turns your background into 30 ideas in a minute.
Step 2: Check demand (search volume and Google Trends)
Demand means real people are typing questions about this topic into Google every month. You check it in two places.
Search volume. Put a few of the main keywords for each niche into a keyword tool and look at monthly volume. You’re not looking for one huge keyword. You’re looking for lots of keywords with a few hundred to a few thousand searches each. In lesson 3 I compare Ahrefs, Semrush, Moz and the free alternatives, so for now use whatever free tool you like, or the free trials I cover in lesson 4.
Here’s a real example. Say “home espresso for beginners” is on your list. These are US monthly volumes from Ahrefs, pulled in November 2026:
| Keyword (example niche: home espresso) | US searches / month | Keyword Difficulty |
|---|---|---|
| best coffee grinder for espresso | 3,600 | 14 |
| how to make espresso at home | 3,400 | 0 |
| best espresso machine for beginners | 1,800 | 10 |
Source: Ahrefs Keywords Explorer, US data, November 2026. This is an illustration of the method, not a niche I run.
That’s healthy demand: three keywords with thousands of searches, plus a buying keyword (“best espresso machine for beginners”) and an informational one (“how to make espresso at home”) that can feed it.
Google Trends. Next, type the main topic into Google Trends and set the range to five years. You want a line that’s flat or rising. A line that spiked once and has been falling since is a fad (more on that in the red flags). Seasonal waves are fine, as long as the peaks aren’t getting smaller every year.
Step 3: Check competition (KD, DR of the top 10, forums ranking)
Demand is pointless if you can’t get onto page one. To check competition, look at three things for your main keywords.
- Keyword Difficulty (KD). Ahrefs scores this from 0 to 100 based on how many sites link to the top-ranking pages. For a brand-new site, I look for plenty of keywords under 20. Even the keyword this lesson targets, “how to find a niche”, sits at KD 2 with 150 US searches a month, which is exactly the kind of gap a new site can take.
- Domain Rating (DR) of the top 10. Google the keyword and check the DR of each ranking site with Ahrefs’ free website authority checker or the toolbar in your SEO tool. If every result is DR 80+ (big magazines, Amazon, government sites), skip it. If you see a few results in the DR 0 to 30 range, a new site can compete.
- Forums and user content ranking. When Reddit threads, Quora answers or old forum posts sit on page one, Google is telling you it can’t find a good, dedicated article for that search. That’s your opening.
This is also why newer topics are so attractive. In lesson 2 I’ll show you why AI and SaaS sub-niches are easier to rank in: most of the sites covering them are young, so they haven’t had years to build up links.
Step 4: Check the money (affiliate programs and CPC)
A niche needs a way to earn before you build anything. Do two quick checks.
Affiliate programs. Google “[your niche] affiliate program” and look through networks like Amazon Associates, Awin, Impact and ClickBank. I list the networks I actually use on my resources page, and lesson 14 goes deep on finding programs. You want at least three or four products or brands you could genuinely recommend, ideally with recurring commissions or decent order values.
Cost per click (CPC). CPC is what advertisers pay Google for one click on that keyword. It’s a shortcut for how much a visitor is worth. Low CPCs across the board usually mean the audience doesn’t buy much. For example, “best espresso machine for beginners” shows a CPC of about $0.45 in Ahrefs, while “best ai writing tools” (2,200 US searches a month, KD 15) shows about $2.50. Both can work, but software buyers are worth more per visitor.
If you want to see what “the money” looks like in practice, my affiliate case study breaks down a site earning $150,000 a year from hosting reviews, and how 3 set-and-forget websites earn me about $2,000 a month shows how small, focused sites can pay too.
Step 5: Check depth with the 50-post test
Open a blank document and try to list 50 article titles for the niche in one hour. Mix the types: how-to guides, “best X for Y” lists, comparisons (“X vs Y”), reviews and beginner explainers.
- Under 30 titles: too narrow. You’ll run out of things to say and Google won’t see you as an authority.
- 30 to 50: workable, especially if you can widen the audience a little.
- 50+ with ideas left over: pass. You have room to build topic clusters, which I teach in lesson 12.
The depth test is also your bridge to the next lessons. Those 50 titles become your first keyword list, and they’ll help you choose a domain name that’s broad enough to grow into.
The niche scorecard
Score each niche on your shortlist from 1 (poor) to 5 (great) on all five checks. Anything under 15 out of 25 gets dropped. If two niches tie, pick the one you’d happily write about for two years.
| Check | Score 1 if… | Score 5 if… |
|---|---|---|
| 1. Knowledge | You’d be learning from zero | You’ve done it, studied it or spend money on it |
| 2. Demand | Few keywords with volume; Trends falling | Many keywords with hundreds to thousands of searches; Trends flat or rising |
| 3. Competition | Top 10 all DR 80+; most KDs over 40 | Low-DR sites and forums on page one; many KDs under 20 |
| 4. Money | No affiliate programs; CPCs near zero | Several programs, good order values or recurring commissions; solid CPCs |
| 5. Depth | Fewer than 30 post ideas | 50+ post ideas with more coming |
| Total | Under 15: drop it · 15 to 19: maybe · 20+: shortlist it | |
Use the same tool and the same country for every niche you score, so the numbers are comparable.
Red flags: niches to avoid as a beginner
Some niches look great on volume and still trap beginners. Watch for these three.
YMYL niches
“Your Money or Your Life” topics, like health, medical, legal and serious personal finance, get extra scrutiny from Google. It wants proof of real expertise and trust, and you’re up against banks, hospitals and big publishers. “best credit cards” gets 256,000 US searches a month, but it has a KD of 41 and is owned by huge financial sites. “best weight loss supplements” (11,000 searches, KD 28) looks easier on paper, but it’s a health claim minefield. Unless you’re qualified in the field, leave these for later.
One-product niches
A site about a single product (one gadget, one app, one brand) lives and dies with it. If the product is discontinued, the affiliate program closes or the brand changes its commission, your income goes with it. Go one level up: not one robot vacuum, but robot vacuums for pet owners.
Fads
Fads have huge volume and very short lives. “fidget spinner” still gets 23,000 US searches a month with a KD of just 3, which looks like a gift until you open Google Trends and see the 2017 spike it has never come close to again. Low difficulty on a dying topic is a trap. You want demand that will still be there when your site is two years old.
You can also buy a site that already has a proven niche and traffic, and grow it. It costs more upfront, but you skip the guessing. I explain how I do it on my buying websites page.
A Claude prompt to brainstorm 30 niches from your background
If your Step 1 list feels thin, let Claude do the brainstorming. Give it your background and the five checks, and ask for ideas you can then test. Copy this into Claude and fill in the brackets.
Claude can’t see live search data unless you connect an SEO tool to it (I show you how in lesson 11), so treat its volume guesses as rough. Take its top ideas and run each one through Steps 2 to 5 with real numbers.
- Write down 20 to 30 raw niche ideas from your work, studies, hobbies and problems, or run the Claude prompt above.
- Cut the list to your best 3 niches.
- Score each of the 3 out of 25 using the niche scorecard, with real volume and KD numbers from a keyword tool.
- Keep the winner (and the runner-up) handy. You’ll check them against AI and SaaS sub-niches in lesson 2.
Want to skip the guesswork? Here are the two ways I can help you get to your first earnings faster.
The One-Pager Strategy
- Build and launch a single monetised page with AI in days, not months
- Low effort: one focused page instead of a 50-post content plan
- Runs alongside your job, no big budget needed
The Full System
- Full video course on running websites as a real business
- A dedicated course ebook to follow along with
- Private student group: share wins, ask advice and collaborate with other students
- Mentorship from me, and you can reach out whenever you’re stuck
- Regular updates in the group on what’s changing in SEO and AI
Compare everything on the courses page.
Why new AI and SaaS sub-niches are easier for a new site to rank in.
Previous← Course homeSummary: how to find a niche
To find a niche, start with 20 to 30 ideas from what you already know, then put each one through five checks: demand (search volume and Google Trends), competition (KD, the DR of the top 10 and whether forums rank), money (affiliate programs and CPC), depth (50 post ideas) and your own knowledge. Score them on the niche scorecard, drop anything under 15 out of 25, and avoid YMYL topics, one-product niches and fads. Once you have a winner, you’re ready for lesson 2, and later for choosing a domain name that fits it.
Get each new lesson in your inbox
Join my email list and I’ll let you know as each lesson goes live. You’ll also get your free copy of The Borrowed Playbook: how to borrow the social media strategies already working for other products and use them to send traffic to your site, with little to no money.
- A heads-up when each new lesson is published
- Real case studies and what’s working right now
- The free Borrowed Playbook guide, sent straight away
You’ll join my newsletter. Unsubscribe anytime.
You’re in. Now check your inbox.
I’ve just sent an email to your inbox. Open it to grab your Borrowed Playbook. If it asks you to confirm your subscription first, click the button and the guide follows right after.
- Can’t see it? Check your Promotions or Spam tab and mark it “not spam”.
- Meanwhile, carry on to lesson 2 and see why AI and SaaS sub-niches are easier to rank in.
Frequently asked questions
How do I find a niche for my website?
List 20 to 30 ideas from your work, studies, hobbies and problems you’ve solved. Then test each one for search demand, beatable competition, affiliate programs to earn from and enough depth for 50 or more posts. Score them out of 25 and pick the highest-scoring niche you’d enjoy writing about for two years.
What is niche research?
Niche research is the process of checking whether a website topic is worth building on before you invest in it. It means looking at search volume, Google Trends, keyword difficulty, the strength of the sites already ranking, the affiliate programs and ad value in the niche, and how many articles the topic can support.
How do I choose a niche for a blog?
Choose a niche that combines a topic you know, a specific audience and a clear way to earn. It should have plenty of keywords with real search volume, a page one that includes some smaller sites or forums, and enough depth for at least 50 posts. Avoid YMYL topics, single-product niches and fads.
How do I find low competition niches?
Look for keywords with low Keyword Difficulty, ideally under 20, and Google them to check the Domain Rating of the top 10 results. If you see low-DR sites, Reddit threads, Quora answers or old forum posts on page one, the niche has gaps a new site can fill. Newer topics, like AI and SaaS sub-niches, often have the weakest competition.
Should I pick a niche I’m passionate about or one that makes money?
Pick one that does both, or at least one you know well and can earn from. Passion alone doesn’t pay if nobody searches for the topic or there’s nothing to sell. Money alone burns out fast if you hate the subject. The niche scorecard weighs knowledge and money equally for that reason.
How many keywords does a niche need?
A rough rule is enough keywords for 50 or more articles, with many of them getting a few hundred to a few thousand searches a month. One big keyword isn’t enough. You want a wide spread of smaller, related searches you can rank for one by one.
Can I change my niche later?
You can widen a niche into closely related topics, which is why a broad, brandable domain name helps. Switching to an unrelated niche usually means starting a new site, because your existing content and links won’t support the new topic. It’s worth spending a few hours testing your niche properly now.
About Eugene Madondo
Eugene is a Cape Town-based SEO specialist and digital entrepreneur who has been earning online since 2012. He has worked in SEO at a tech startup, an agency and in-house tech roles, ran a consulting business, and has generated over $1.4 million from his own affiliate websites since 2020. He teaches website operators through Rank to Earn.
Back to the free website course overview, or see the tools I use on my resources page.