Japan Ecommerce Pricing Models: Platform Fees vs Agency Cost

Ask three Japan e-commerce partners for a quote and you will typically get three numbers that cannot be compared: a flat monthly retainer, a percentage of sales, and a "base plus commission" hybrid. Behind them sit a second set of costs entirely — the fees the marketplace itself takes before anyone touches your storefront. Foreign brands routinely conflate the two, compare on the headline figure, and discover months later that the "cheaper" option excluded the things that turned out to cost the most. This article separates the layers so you can actually budget.

There are two questions, and they need to stay separate. First: what does the platform charge me to sell here — the unavoidable cost of being on Rakuten, Amazon Japan, Yahoo! Shopping, or running your own Shopify store? Second: what does it cost to have someone operate that store for me, and how do I read a proposal? Everything below is organized around that split, because the most expensive mistakes come from treating one as if it included the other.

Layer 1: What Each Platform Charges to Sell

Every marketplace prices differently, and the differences are structural, not cosmetic. One charges a fixed monthly fee, one charges per-category commission, one charges nothing to open but takes its cut through the point system, and one charges you nothing to sell but gives you no traffic. Knowing the shape of each is the prerequisite for reading any agency quote, because these are the costs that sit outside whatever you pay an operator.

Rakuten Ichiba — monthly plan plus a percentage

Rakuten runs on a store-plan model. There are three plans — the entry Ganbare! plan at ¥25,000/month (billed as twelve months up front, so roughly ¥300,000/year), the Standard plan at ¥65,000/month, and the Mega Shop plan at ¥130,000/month — plus a one-time registration fee around ¥60,000, all before tax. On top of the fixed fee sits a percentage system usage fee, which Japanese operator guides consistently report in the region of 2–7% depending on plan and revenue tier (the cheaper the fixed plan, the higher the percentage). Then there is Rakuten Pay settlement at roughly 2.5–3.5% of sales and point-program funding that starts around 1% and rises with campaign participation. Add it up and Japanese guides commonly cite total platform overhead in the region of 10–15% of sales once every variable cost is counted — before you spend a yen on advertising or agency support. We break down the plan-choice math in more depth in our Rakuten agency guide.

Amazon Japan — per-account plus category referral fees

Amazon Japan is the most familiar model for most foreign brands. You pay either ¥100 per item sold (Individual plan) or a flat ¥4,900/month (Professional plan), and on every sale a category referral fee that generally falls between 5% and about 15%, varying by product category. If you use Amazon's fulfilment network, FBA fees for storage and shipping sit on top of that. One timing note worth building into a 2026 budget: Amazon Japan is updating referral fees, FBA fulfilment fees, and aged-inventory surcharges from April 2026 — the changes are incremental (an electronic-accessory item at ¥1,000 moves from a 10% to a 10.4% referral fee, for example) but they move the baseline upward, so use current published rates rather than older guides. Our Rakuten versus Amazon Japan comparison covers how the two platforms' economics differ beyond the headline fees.

Yahoo! Shopping — free to open through August 2026, paid through points

Yahoo! Shopping is the outlier: through 31 August 2026, initial fee, monthly fee, and sales royalty are all ¥0. That makes it the cheapest platform to open on paper, and the reason many guides recommend it for low-fixed-cost testing. But "no fixed fee" is not "free" — sellers still carry variable costs when a sale happens: payment processing, campaign participation, and above all PayPay and point funding (starting around 1% and rising with the campaigns you opt into). Japanese operator guides put the realistic all-in variable cost in the region of 5–8% of sales. The trade is straightforward: you avoid the fixed monthly commitment, but the platform's economics push you toward funding points to get visibility. From September 2026, Yahoo! adds a monthly system usage fee of ¥10,000 (excl. tax) and a 2.5% sales royalty — see our outsourced Yahoo! Shopping store management guide for the full change and the revised arithmetic. Our Yahoo! Shopping versus Rakuten comparison goes deeper on when the free-to-open model actually pays off.

Shopify — no marketplace fee, but no marketplace traffic

Shopify is a different animal because it is not a marketplace at all — it is your own store, so there is no referral or system-usage fee taking a cut of each sale. Instead you pay a monthly software plan (Basic from around $39/month billed monthly, the mid "Shopify" plan around $105/month, Advanced around $399/month, and Shopify Plus positioned near ¥368,000/month in Japan) plus payment processing — Shopify Payments takes roughly 2.9% + a fixed fee per card transaction. There is one Japan-specific catch that generic pricing guides miss: a large share of Japanese shoppers pay by konbini (convenience store), bank transfer, or cash on delivery, which run through third-party gateways, and using a third-party gateway triggers Shopify's extra transaction fee (2% on the Basic plan, lower on higher plans). So your effective Shopify cost in Japan depends heavily on your payment mix, not just your plan. And because there is no marketplace behind you, the real cost of Shopify is the traffic you have to buy or earn yourself — a point we cover in Rakuten versus Shopify for foreign brands, or in our Amazon-first sequencing guide if Amazon Japan is the marketplace you're weighing it against instead.

The platform fees at a glance

All figures are before tax and are order-of-magnitude references drawn from published platform pages and Japanese operator guides, not quotes; percentage bands vary by plan, category, and revenue tier. The point of the table is not the exact number — it is that the four platforms are not priced on the same axis, so "which is cheapest" has no answer until you model it against your own product category, price point, and payment mix.

Layer 2: The Three Agency Pricing Models

Once you know the platform costs, the second question is what it costs to have someone operate the store. Japanese operation-agency (運営代行, unei daiko) guides describe three pricing structures, and almost every proposal you receive will be one of them or a blend. Knowing the shape lets you read a quote instead of just reacting to the number.

• Fixed monthly retainer (月額固定型). A flat fee regardless of sales. Japanese guides cite roughly ¥200,000–500,000/month for full store operation, dropping to around ¥50,000/month for consulting-only engagements and ¥50,000–100,000 for partial task outsourcing. The advantage is predictability and budget clarity; the risk is that you pay the same in a flat month as in a strong one, and the agency has no direct upside from growing your sales.
• Commission / revenue share (成果報酬型・レベニューシェア). A percentage of sales, generally cited at 5–20%, and often 10–20% for marketplace (mall) operation specifically. The advantage is alignment — the agency earns more only when you sell more, which lowers your risk at low volume. The risk is the mirror image: in a strong month you write a large cheque for work that did not necessarily scale in proportion, and the percentage can quietly exceed a fixed retainer once you cross a certain revenue level.
• Hybrid (複合型). A smaller fixed base plus a percentage — the most common structure for serious full-operation mandates. Guides commonly describe a ¥200,000–300,000 base plus roughly 5–15% of sales, and for mall operation specifically a pattern of around 10–20% of sales plus a ¥50,000–150,000 fixed fee. It splits the risk: the base covers the agency's fixed operating cost, and the percentage keeps incentives aligned. Most established operators land here.

Setup or onboarding fees — store build, initial catalog, first campaign plan — are almost always quoted separately from any of these ongoing models, and they are one of the line items brands forget to ask about. Comprehensive full-operation mandates that include page production, advertising, and customer service run materially higher than a partial-task retainer, so a single "how much does an agency cost" number is meaningless without a scope attached.

Why Three Quotes Never Match — and How to Read Them

Here is the scenario that sends brands to us most often: they have three proposals in hand — say a ¥250,000 flat retainer, a 12% revenue share, and a "¥150,000 + 8%" hybrid — and no way to tell which is actually cheaper, because they are priced on three different axes. The instinct is to pick the lowest headline number. That is exactly the wrong move. Two disciplines turn incomparable quotes into a real comparison.

Model every quote against your own realistic sales

A fixed retainer and a percentage deal cross over at a specific revenue level, and which one is cheaper flips depending on where your sales actually land. A 12% revenue share is cheaper than a ¥250,000 retainer only up to about ¥2,000,000/month in sales; above that, the percentage costs more. So the honest comparison is not "12% versus ¥250,000" in the abstract — it is "12% versus ¥250,000 at the ¥1.5M/month I realistically expect in year one, and again at the ¥4M I hope for in year two." Run each quote through your own low, expected, and optimistic sales figures. The ranking often changes across those three scenarios, and that tells you as much about your risk appetite as about the agencies.

Normalize the scope before you compare the price

The number that gets glossed over in every proposal is what is excluded. Before comparing two quotes, make each agency write down, line by line, exactly what is included and what is not — and pay special attention to these items, which routinely migrate between "included" and "your cost" from one proposal to the next:

• Platform fees — Rakuten's plan and system fees, Amazon's referral and FBA fees, Yahoo's point funding. These are almost always the brand's cost, but a vague quote can imply otherwise.
• Advertising spend — the media budget itself (RPP, sponsored products) versus the fee to manage it. Some quotes bundle management into the retainer; some charge a percentage of ad spend on top.
• Point-campaign and coupon funding — on Rakuten and Yahoo this is a real, recurring cost, and it is the brand's money, not the agency's.
• Photography and creative production — often a separate line, sometimes assumed to be included.
• Regulated-category compliance — if you sell cosmetics, supplements, or anything touching 薬機法, the compliance work has a cost and someone has to own it. Our market-entry cost and timeline guide maps where these regulated-category costs land in the wider budget.

Once every quote is normalized to the same scope and modelled against the same sales figures, the "cheapest" one frequently turns out to be the most expensive — a thin scope over a low headline rate is the most common way a Japan EC budget blows out in month four.

A Simple Framework for Choosing a Pricing Model

Setting aside which agency, the structure that suits you depends mostly on your stage and your risk appetite:

• Early and uncertain about volume? A revenue-share or hybrid-leaning-percentage model lowers your downside — you are not paying a heavy fixed fee before you know whether the channel works. The trade is that you give up more of the upside if it takes off fast.
• Established, predictable sales, and want cost certainty? A fixed retainer (or hybrid with a low percentage) caps your cost and keeps a strong month from generating a surprise invoice. The trade is weaker built-in incentive for the agency to push growth, which you offset with clear performance targets in the contract.
• Somewhere in between, or running multiple channels? A hybrid is popular precisely because it hedges — a base that keeps the lights on plus a percentage that keeps incentives aligned. Just make sure the base is sized to the actual operating load, not padded.

Whatever the structure, insist on a scope written line by line and reporting you can actually read. The framework for judging the agency itself — track record, native Japanese capability, foreign-brand experience — is a separate question we lay out in how to choose a Japan EC agency. Pricing is only half the decision; a low rate over a weak operator is the worst outcome of all.

Common Pricing Traps for Foreign Brands

The mistakes cluster in predictable places. Comparing quotes on the headline number without modelling them against realistic sales, so the "cheap" percentage deal becomes the expensive one at scale. Assuming an agency fee includes platform fees, ad spend, or point funding when it almost never does. Forgetting the separately-quoted setup fee entirely. Choosing the platform purely on lowest fixed cost — Yahoo! Shopping is "free" to open, but a store that funds no points gets no visibility, so the real cost shows up on the variable side. And treating price as the whole decision when scope and operator quality determine whether you get value at any price. Each of these is a version of the same error: reading one number instead of the whole cost structure.

How LAUNOVA Helps

LAUNOVA works exclusively with foreign brands entering the Japanese e-commerce market, across Rakuten Ichiba, Amazon Japan, Yahoo! Shopping, and Shopify — which means part of the job is simply helping a brand read the true cost structure of the channel it is considering before committing to anyone. Our Japan EC operation service runs stores on an ongoing basis, and our market-entry partnership covers the earlier decisions about which platform and cost model fit your product and stage. Because a first-time entrant, a single-channel operation, and a full multi-channel mandate are genuinely different jobs, we quote each engagement per brand by scope rather than selling a fixed package — the scope drives the number, not the other way around. If you are holding quotes you cannot compare, or trying to budget a Japan launch, tell us your channels and stage and we'll help you read the real cost →

FAQ

Q: What are the typical pricing models for a Japan ecommerce agency?
Japanese operation-agency (運営代行) guides describe three structures. A fixed monthly retainer — commonly cited around ¥200,000–500,000 for full store operation, less for partial or consulting-only mandates — is predictable but charges the same whether sales move or not. A commission or revenue-share model — generally 5–20% of sales, and often 10–20% for marketplace (mall) operation — aligns the agency with growth but produces a large bill in a strong month. A hybrid model — a smaller fixed fee plus a percentage, frequently a ¥200,000–300,000 base plus roughly 5–15% — is the most common structure for serious full-operation mandates. Setup or onboarding fees are usually quoted separately. These are order-of-magnitude industry references, not quotes; the real number tracks scope.

Q: How do I compare three agency quotes that use different pricing structures?
You cannot compare them on the headline number, because a ¥250,000 fixed retainer, a 12% revenue share, and a "¥150,000 + 8%" hybrid are only comparable once you translate them onto the same basis. Do two things. First, model each quote against your own realistic monthly sales — a percentage deal and a fixed deal cross over at a specific revenue level, and which is cheaper flips depending on where you land. Second, normalize the scope: make every agency write down exactly what is included and, more importantly, what is excluded — platform plan fees, ad spend, point-campaign funding, photography, and regulated-category compliance are the line items that quietly move between "included" and "your cost" from one proposal to the next. A cheaper rate over a thinner scope is usually more expensive in total.

Q: How much do the Japanese marketplaces themselves charge to sell?
Each platform charges differently. Rakuten Ichiba has a monthly store-plan fee (¥25,000–130,000 depending on plan, before tax) plus a percentage system usage fee and point-program funding. Amazon Japan charges ¥4,900/month for a Professional account (or ¥100 per item sold on the Individual plan) plus category referral fees generally in the 5–15% range, with FBA fulfilment fees on top. Through 31 August 2026, Yahoo! Shopping has no initial or monthly fee at all, but sellers still carry payment processing and PayPay/point funding that typically total roughly 5–8% of sales; from September 2026, Yahoo! adds a monthly system usage fee of ¥10,000 (excl. tax) and a 2.5% sales royalty. Shopify has no marketplace fee because it is your own store — you pay a monthly software plan (from around $39/month) plus payment processing, and you buy all your own traffic. These are platform costs that sit outside any agency fee.

Q: Does an agency fee include the platform fees, or are those separate?
They are separate, and this is the single most common source of confusion when comparing quotes. An agency operates within the platform costs; it does not absorb them. Rakuten's store-plan and system-usage fees, Amazon's referral and FBA fees, Yahoo's point funding, ad spend, and point-campaign budgets are the brand's costs regardless of who runs the store. A proposal that is vague about which side of that line a cost falls on should be pushed on before you sign — the gap between a "10% all-in" quote and a "10% plus you fund platform fees and ads" quote is enormous.

Q: What does LAUNOVA charge?
Because a first-time market entrant, a brand that needs one channel operated, and an established seller who wants full multi-channel operation are genuinely different jobs, LAUNOVA quotes each engagement per brand by scope rather than publishing a fixed package price. The right way to get a number is to describe your channels, your current stage, and where your capability gaps are; the scope drives the price, not the other way around. Tell us where you are and we will scope it.