When people start thinking about affiliate marketing, the first instinct is often to ask: “What products sell the most?”
Cosmetics? Electronics? Food? Expensive chairs? Travel? Financial services?
But that skips a more important question:
Why would this reader buy this particular thing online?
A product that sells well in society, a product that is naturally suited to e-commerce, and a product that generates attractive affiliate revenue are not the same thing.
Milk sells every day. That does not mean someone reading “17 Best Bottles of Ordinary Milk” wants to order one carton online.
Water, rice, pet food, or replacement brush heads are different. Suddenly there are strong online reasons: heavy to carry, hard to find locally, exact model required, repeatedly purchased.
So the real unit of analysis is not the product name.
It is article context × reason to buy online × monetization economics.
0. Start with online necessity, not generic popularity
A practical affiliate filter looks roughly like this:
article relevance × e-commerce reason × purchase intent × expected payout
If one factor is near zero, the opportunity weakens sharply.
Putting luxury perfume into an article about shoulder stiffness does not become logical because the commission is 20%.
By contrast, odor-control products in an article about cat litter make sense. The reader already has a problem, and the product is a plausible next action.
The goal is not to turn every article into a catalog.
An affiliate link should be one possible next step for the reader, not the reason the article exists.
1. What actually sells online?
Japan's Ministry of Economy, Trade and Industry estimated the 2024 domestic B2C e-commerce market at ¥26.1 trillion, with physical-goods e-commerce at ¥15.2194 trillion.[1]
Major physical categories included:
- Food, beverages and alcohol: ¥3.1163 trillion; EC ratio 4.52%
- Clothing and apparel: ¥2.7980 trillion; 23.38%
- Home appliances, AV equipment and PCs: ¥2.7443 trillion; 43.03%
- Household goods, furniture and interiors: ¥2.5616 trillion; 32.58%
- Books, video and music software: ¥1.8708 trillion; 56.45%
- Cosmetics and pharmaceuticals: ¥1.0150 trillion; 8.82% [1]
The interesting part is that market size and e-commerce suitability are different variables.
Food has a huge online market but a low EC ratio. Books and packaged media have a smaller total online market, yet more than half of their relevant commerce is online.
Travel services alone reached ¥3.5249 trillion.[1]
So the useful question is not only “Do people spend a lot on this?” but also:
Is buying this category online already normal behavior?
2. Why do people choose online instead of a store?
A 2025 Japanese survey found that “lower price” was selected by more than 60% as a situation favoring online shopping, while more than 40% cited being able to use accumulated points.[2]
Another 2025 survey found that assortment, free shipping and lower price were leading reasons for choosing an e-commerce mall. Repeat purchases were especially visible for consumables such as daily necessities, beauty products and beverages.[3]
A cosmetics survey similarly found price, convenience of buying from home, reviews and delivery among the leading reasons to buy through e-commerce.[4]
A meta-analysis of e-commerce adoption research also found that attitude and convenience were among the strongest drivers.[5]
Translated into ordinary human language, online buying becomes attractive when it is:
- cheaper,
- rewarded with points or discounts,
- easier than visiting a store,
- delivered instead of carried,
- unavailable nearby,
- offered in more variants,
- easier to compare,
- supported by reviews,
- safer to buy from an official source,
- repeatedly replenished,
- available 24/7,
- or more private.
Rarity is only one motive.
An utterly ordinary product can become an excellent e-commerce product simply because carrying it home is annoying.
Six two-liter bottles can convert almost anyone into a logistics enthusiast.
3. What sells most is not necessarily what earns most
A simplified affiliate revenue equation is:
traffic × link CTR × conversion rate × order value × commission rate × approval rate
That is why expensive does not automatically mean profitable.
Amazon Japan's standard schedule currently assigns different rates by category: for example, 8% for food and some fashion categories, 5% for beauty and pet supplies, and 2% for many electronics and PC categories.[6]
Rakuten Marketplace affiliate rates are generally 2–4%, with attribution rules and per-item payout limits.[7]
Direct advertiser programs can look very different. In A8.net's September 2026 listings, examples included 10% on smart-home products, 15% on some beauty products and ¥20,000 for a chair purchase, while its high-payout list included tens of thousands of yen for broadband activation, course enrollment or specified financial conversions.[8][9]
But A8's program ranking is not a sales ranking.
Confusing the two produces a wonderful monster:
“This payout is ¥200,000. Clearly every article should now lead to hair restoration.”
A high payout tells you something about what acquiring a customer may be worth to the advertiser. It does not prove that your reader has any reason to become that customer.
4. Search goods and experience goods explain a lot
Consumer research often distinguishes between “search goods,” whose important attributes can largely be evaluated before purchase, and “experience goods,” whose quality becomes clearer only after use.[10]
PCs and many electronics lean toward search goods.
Processor, memory, dimensions, power usage, weight and price can all be compared.
Perfume, food, cosmetics and bedding contain stronger experience elements.
A web page can say a fragrance is “fresh,” but HTTP still cannot transmit smell.
That is why physical stores remain valuable.
However, online information can reduce the unknown part of an experience good.
Reviews, texture descriptions, sizing notes, return policies, ingredients, measurements, photos, comparisons and “good for / not good for” guidance can move a buyer from “I know nothing until I try it” to “I can at least avoid obvious mismatches.”
That is exactly where an article can create value.
5. Good affiliate content reduces uncertainty more than it pushes products
A 2024 meta-analysis combined 156 studies and 69,006 observations and found meaningful relationships between online-review factors and purchase intention; review valence showed the strongest association among the examined antecedents.[11]
So a useful article is not one that repeats “Amazing! Buy now!” twenty times.
Readers usually need answers such as:
- Who is this for?
- Who should skip it?
- What is actually different?
- What are the weaknesses?
- Why do negative reviewers complain?
- Is there a cheaper substitute?
- Should I try it in a store first?
- Is Amazon, Rakuten or the official shop the logical place to buy?
- Do I need it now at all?
The strongest affiliate article behaves more like a procurement analyst than a salesperson.
Reducing uncertainty may sometimes lead to “do not buy this.”
That can still create trust that converts elsewhere later.
6. Does a “Top 20” really need twenty products?
More options are not always bad.
But a meta-analysis of choice overload found that overload becomes more likely when the choice set is complex, the decision task is difficult, preferences are uncertain, or the decision-maker mainly wants to minimize effort.[12]
So a “30 Best Products” page can accidentally take the confusion a reader brought from search and multiply it inside the article.
The value is not the number of products.
It is the compression of decision rules.
For example:
- lowest cost → A
- lowest effort → B
- longest-term use → C
Now the reader can map a real priority to a smaller set.
The real protagonist of comparison content is not the product list.
It is the branching rule that tells different readers when to choose what.
7. Why cosmetics, telecom, finance and education appear so often in affiliate programs
It is not necessarily because they are the world's most-purchased categories.
Advertisers can pay more to acquire customers when the underlying economics support it:
gross margin × retention × customer lifetime value × value of a new acquisition
Cosmetics and consumables can repeat.
Telecom contracts can persist.
Education can have high transaction value.
In finance, recruiting or other services, one qualified new customer or consultation can itself be economically valuable.
So the categories most visible in affiliate marketing partly reflect customer-acquisition economics, not a universal ranking of consumer purchases.
Yano Research estimated Japan's FY2025 affiliate market at ¥459.8 billion and noted continued strength in e-commerce-mall models and selected segments, while also pointing to changes in SEO traffic and stricter KPI evaluation.[13]
Finding a large commission is not the same as finding a gold mine.
A high-paying offer without contextual demand is simply an oil field on another continent.
8. For implementation, classify “reasons to buy online” before products
If products are going to be attached article by article, four filters are useful:
| Axis | Question |
|---|---|
| Article relevance | Is the offer a natural continuation of the reader's current problem or goal? |
| E-commerce reason | Is there a reason to buy online: price, weight, scarcity, comparison, reviews, delivery, official authenticity? |
| Purchase intent | Is the reader merely learning, actively comparing, or close to purchase? |
| Economics | What are the price, rate, CVR, approval rate, repeat potential and program durability? |
A sensible order is:
relevance → online-buying reason → purchase intent → economics
Starting with economics makes the same lucrative offer mysteriously grow out of every article.
Starting with relevance also allows the correct answer: “No affiliate link belongs here.”
The target is not simply “a product that sells.”
It is:
a product for which you can explain why this reader, after this article, would reasonably buy it online now.
At that point, the link can become a useful next action rather than decorative advertising.
In Japan, stealth marketing has been regulated as misleading representation under the Act against Unjustifiable Premiums and Misleading Representations since October 1, 2023. Affiliate disclosures must be clear enough that readers understand the advertising nature of the content.[14]
Place products naturally.
Do not hide advertising naturally.
Those are separate ideas.
References (14)
- Ministry of Economy, Trade and Industry, Japan (2025), FY2024 E-Commerce Market Survey meti.go.jp
- MyVoice Communications (2025), Online Shopping Survey, 21st wave myvoice.co.jp
- Cross Marketing Group (2025), E-commerce Usage Survey prtimes.jp
- eBay Japan / Qoo10 (2025), Cosmetics purchasing and review survey prtimes.jp
- Zerbini C, Bijmolt THA, Maestripieri S, Luceri B. (2022). Drivers of consumer adoption of e-Commerce: A meta-analysis doi.org
- Amazon Associates Japan, Standard Commission Income Statement affiliate.amazon.co.jp
- Rakuten Affiliate Guidelines affiliate.rakuten.co.jp
- A8.net, High-payout program ranking support.a8.net
- A8.net, Program rankings support.a8.net
- Schmid B, Axhausen KW. (2019). In-store or online shopping of search and experience goods doi.org
- Qiu K, Zhang L. (2024). How online reviews affect purchase intention: A meta-analysis across contextual and cultural factors doi.org
- Chernev A, Böckenholt U, Goodman J. (2015). Choice overload: A conceptual review and meta-analysis doi.org
- Yano Research Institute (2026), Affiliate Market Survey yano.co.jp
- Consumer Affairs Agency, Japan, Stealth marketing regulation caa.go.jp
