How to Choose the First Topic for a New Affiliate Site (A 30-Minute Decision)
The first topic is the decision new affiliate site owners spend the longest on and get the least help with. The usual advice is "find a profitable niche", which is true and unusable.
What actually decides whether a first topic works is five things you can check in half an hour with a browser, and a topic that passes all five is a good first topic whether or not it appears on anyone's list of profitable niches.
This post is the thirty minutes: the five questions, how to check each, a scoring sheet for three candidates, what to do with a tie, and the mistakes that look fine at post five and show at post fifty. It sits before the niches post, which is where candidates come from; this is how to choose between them.
The short answer
Write down three candidate topics. For each, check: Amazon sells it at $40 or more with real review counts; "best [topic] for" produces a long list of specific searches; the category has 30 to 60 buying keywords, not 300; you could name the product you'd buy and why without research; and the results for its long-tail keywords include small sites and forums.
Score each 0 to 2. Highest total wins; a tie goes to the one you know better. Start writing the same day.
The five questions
| Question | How to check (5 minutes) | 2 points | 0 points |
|---|---|---|---|
| 1. Do people buy it on Amazon, at a price worth a commission? | Search the category on Amazon; note the top ten's prices and review counts | Typical price $40+, thousands of reviews | Under $20, or few reviews, or mostly a different retailer |
| 2. Is there a long tail of buying searches? | Type "best [topic] for" and "[topic] under" into Google; count autocomplete suggestions | Ten-plus distinct "for"/"under" variants | Two or three, all the same thing |
| 3. Can you cover it in 30 to 60 posts? | List the guide keywords from question 2 plus comparisons and questions | The map fits a quarter or two | Hundreds of keywords (too broad), or under fifteen (too narrow) |
| 4. Do you know enough to have an opinion? | Without looking anything up, name the product you'd buy and the one most people overrate | Both, with a reason | Neither |
| 5. Are the long-tail results beatable? | Search two "best [topic] for [situation]" keywords; look at positions 4 to 10 | Small sites, forums, thin pages | All national publications and retailers |
1. Bought on Amazon, at a price
Commission is a percentage of price, so a category where the typical product is $15 pays a third of one where it's $45, for the same work. Review counts in the thousands mean the products sell on Amazon specifically, rather than at a specialist retailer.
Categories that fail this are the ones that look fun and never pay: cheap consumables, products people buy in person, things Amazon doesn't stock well.
2. The long tail
A first site can't win "best bike lights". It wins "best bike lights for commuting", "for mountain biking", "under $30", "for night riding", and the test is whether those exist. Autocomplete shows them in ten seconds. A topic with a head term and no tail is a topic for an established site; a topic with twenty tails is a first topic.
3. Coverable in one or two quarters
The whole point of a first topic is to cover it completely, because coverage is what a new site has instead of authority. "Cycling" has thousands of keywords and can't be covered; "bike lights" has forty and can.
If the list from question 2 plus the obvious comparisons and questions runs to 30 to 60, it's the right size; the topical authority post covers why that number matters.
4. An opinion
You'll write the verdict on fifty guides. If you can say, now, which product you'd buy and which one people overpay for, every verdict has a point of view. If you can't, the guides are listings with sentences, and the editing pass that makes a site worth reading has nothing to add. This isn't about passion; competence and a few opinions are enough.
5. Beatable results
Search two of the long-tail keywords and look past the top three at positions four to ten. Forums, Reddit threads, small blogs and thin pages are the sign a new site can get in; a page one that's entirely retailers and national publications is a sign to pick a different topic or a narrower one.
The scoring sheet
| Candidate | Q1 price | Q2 tail | Q3 size | Q4 opinion | Q5 beatable | Total |
|---|---|---|---|---|---|---|
| Bike lights | 2 | 2 | 2 | 2 | 2 | 10 |
| Road bikes | 2 | 2 | 0 (too broad) | 1 | 0 | 5 |
| Bike bells | 0 ($12) | 1 | 0 (too narrow) | 2 | 2 | 5 |
Ten points is rare and anything over seven is a usable first topic. A tie goes to the one you'd rather write fifty posts about. A candidate with a zero on question 1 or question 5 is out regardless of its total, because no amount of writing fixes a category that doesn't pay or a page one you can't reach.
Mistakes that take a year to show
- Starting with the hobby instead of a category. "Cycling" at post 50 is 5% coverage of ten topics; "bike lights" at post 50 is done.
- Chasing the commission rate table. A 10% category with nothing you can say about it loses to a 4% category you know.
- Picking by search volume. Volume lives on head terms a new site won't rank for. The tail is where the first year's traffic is.
- Buying a domain before choosing. The name then boxes the topic in. Choose the topic, then the name.
- Deliberating for a month. The sheet takes thirty minutes and the answer rarely changes with more thought. The cost of a wrong first topic is a month; the cost of a late start is the month plus the deliberation.
After the decision
Same day: map the topic into guides, comparisons, reviews and questions (the keywords from question 2 are the start), pick the two easiest guides, and write them. The content calendar post takes it from there.
Revisit the decision at post fifteen with Search Console open: impressions on most guides means keep going; nothing on any of them after two months means the next candidate on the sheet.
Doing it with AffSERP
Questions 2 and 3 are what AffSERP's Keyword Explorer answers in a minute per candidate: it expands a topic into its "for", "under" and "vs" tails and shows how many there are, which is the size check.
The Competitor Sitemap Explorer shows how much of that list a ranking competitor has covered, which is a proxy for question 5. Questions 1 and 4 are yours: Amazon's prices and review counts, and whether you have anything to say.
FAQ
How do I pick a niche for an affiliate site?
Score three candidate topics on five questions: do people buy it on Amazon at a decent price, does it have a long tail of specific buying searches, can you cover it in about forty posts, do you know enough to have an opinion, and are the results pages for its long-tail keywords beatable? The topic that scores highest in thirty minutes is almost always the right one; a week of deliberation rarely changes it.
Should I pick a niche I'm passionate about?
Pick one you know enough about to have opinions, which is a lower bar than passion and a more useful one. You'll write fifty guides on it; knowing which product is overrated matters, loving the hobby doesn't. Plenty of good sites are run by people who are merely competent in the topic.
How big should the first topic be?
Small enough to cover completely in one or two quarters: a product category with 30 to 60 buying keywords, not a whole hobby. 'Bike lights' is a first topic; 'cycling' is a site with several topics, and starting there means covering none of them. Grow into the neighbours once the first is done.
What if the topic I pick turns out wrong?
You'll know by post fifteen or so: no impressions on any guide, or products that don't convert. Stop, keep the posts, pick the next-best candidate from the sheet and start that cluster. A wrong first topic costs a month; a year spent avoiding the decision costs a year.
Keep reading
