Is one agency across all Japan marketplaces better than a specialist for each platform?

Neither is better in the abstract, and the usual argument — that a single agency cannot match a specialist's depth — is aimed at the wrong variable. The decision that actually costs money is who owns the work that sits between channels, because some of it is genuinely shared. Your price on Rakuten is an input to your Amazon Japan Featured Offer eligibility; your discount history on one channel governs what you are permitted to display as a reference price in the next campaign; one stock pool serves every promotional calendar you run. A specialist per platform can be the right answer, but only if you have decided in advance who arbitrates those shared items, and that person is usually someone on your side of the contract. If nobody has been named, the split model does not have two owners — it has none.

How much does it actually cost to manage two or three separate Japan agencies?

The visible cost is that you pay two or three retainers instead of one, and each will be smaller than a full multi-channel fee but the total is usually higher than a single combined engagement. The cost that is not on any invoice is your own coordination time. Our own working estimate — this is LAUNOVA's planning figure, not an industry benchmark — is roughly two to four hours per additional vendor relationship per month for routine running (separate reporting calls, reconciling numbers that are cut differently, relaying decisions), rising steeply in any month with a major event, because the promotional calendar has to be aligned across parties who do not talk to each other. Two specialists rather than one generalist is therefore not a like-for-like swap; it converts an agency's internal coordination into your coordination, and that work has to land on a named person with the authority to overrule a vendor.

Can my Rakuten price affect my Amazon Japan listing?

Yes, and this is the seam that surprises brands most often. Amazon's own seller-facing guidance states that Featured Offer prices are commonly at or below the lowest priced alternatives, and its Pricing Health tooling advises sellers to bring prices at or below the competitive external price — external meaning outside Amazon, which includes your own storefront on another marketplace. So a deep discount run by a Rakuten specialist is not a Rakuten-only decision; it changes the external reference against which your Amazon offer is measured. We could not verify an amazon.co.jp-published page carrying this wording, so treat the mechanism as Amazon's stated global seller guidance rather than a confirmed Amazon Japan policy statement, and check your own Pricing Health dashboard for how your listings are actually being assessed. Either way, the governance point stands: if two agencies each hold pricing authority on their own channel, neither of them owns this interaction.

Why does a discount on one channel limit what I can run later?

Because in Japan the legality of a discount display depends on your own price history, not just on the discount. The Consumer Affairs Agency guidance on price representations treats a past price as a usable comparison price when, looking back over the eight weeks before the sale, the product was sold at that price for the majority of the period it was on sale — and it is not usable if that price ran for less than two weeks in total, or if more than two weeks have passed since it was last sold at that price. Rakuten operationalises this in its own published Super SALE Search application conditions, which require a track record at the original price for at least four of the preceding eight weeks and within the last two weeks. The practical consequence is that your reference price is a resource that gets consumed: every week spent at a promotional price is a week not spent building the baseline the next campaign needs. Two agencies discounting independently can each be individually reasonable and still leave you ineligible for the event you actually cared about.

At what point should I move from one agency to platform specialists?

The trigger is rarely revenue on its own — it is whether a channel has become big enough and different enough to need decisions made at a depth a shared team cannot reach, and whether you now have someone in-house who can arbitrate between vendors. In practice that means: one channel is carrying enough of your Japan revenue that its advertising alone justifies dedicated attention; you have a named person, on your payroll, who owns pricing and the promotional calendar across all channels and can tell a vendor no; and you are prepared to pay the coordination overhead as a real line item rather than absorbing it invisibly. If the first condition is true but the second is not, splitting usually makes things worse rather than better, because the depth you gain on one channel is paid for with an unowned seam everywhere else. LAUNOVA scopes multi-channel engagements to the modules you actually hand over rather than a fixed package — tell us which channels are live, what your current split looks like and where the handoffs are failing, and we will give you a real answer, including when the honest one is that you do not need us on every channel.

Japan Marketplace Agency: One for All Channels or One Each?

The question arrives at a predictable moment. You launched on one Japanese marketplace, it worked well enough to justify a second, and now a third is under discussion. Your incumbent agency says it can take the new channel. A specialist agency for that channel says the incumbent will run it as an afterthought. Both are describing something real, and neither is describing the thing that will actually cost you money over the next two years.

The pitch on both sides is about depth. The specialist argues that Rakuten, Amazon Japan and Yahoo! Shopping have different advertising logic, different search behaviour and different promotional calendars, so a team that claims all three is probably shallow in at least two. The generalist argues that one partner means one contract, one report and one throat to choke. Our own guide to choosing a Japan EC agency makes the generalist version of this case under platform coverage, and it holds up: a single-platform specialist does leave you shopping for another agency the moment you add a channel.

But depth is a question about how well each channel is run in isolation, and it is not where multi-channel operations break. They break at the joins. Some of the work in a Japanese marketplace operation is genuinely shared between channels — not merely similar across them, but the same object touched from two places. Nobody's platform expertise covers it, because it does not belong to a platform. This article is about those joins, what they cost to leave unowned, and how to decide which purchasing model puts them in competent hands.

The Question Is Usually Argued on the Wrong Axis

Start by separating two things that get bundled together.

Execution depth is per-channel: does whoever runs your Rakuten store understand RPP bidding, the event calendar and how the search algorithm treats your product copy? This is a real question, it is answerable by reference-checking, and a competent multi-channel agency answers it by having platform-specific people rather than one generalist team stretched across three consoles. It is worth noting that this is a question you should ask either model — a "specialist" agency with two Rakuten people is not automatically deeper than a multi-channel agency with six.

Seam ownership is cross-channel: who decides what happens when running channel A well is in tension with running channel B well? This question has no platform expert, because the answer is not a platform fact. It is a governance fact, and the models differ sharply on it. A single agency owns the seam by default — badly or well, but someone inside one organisation has to reconcile it. Two specialists own their own sides and nothing in between, which means the seam falls to you whether or not you noticed you had accepted it.

When a split gets regretted, it is almost always for seam reasons, not depth reasons. The specialist genuinely was better at its channel. That was not the problem.

Three Seams That Belong to Neither Specialist

These are not hypothetical coordination frictions. Each one is a mechanism where an action taken correctly on one channel changes what is possible on another.

1. Your price on one marketplace is an input to your standing on another

Amazon's own seller-facing guidance is explicit that price competitiveness governs Featured Offer eligibility, and that the comparison is not only against other Amazon sellers. Amazon states that Featured Offer prices are commonly at or below the lowest priced alternatives, and its Pricing Health tooling directs sellers to bring prices at or below the competitive external price — external meaning outside Amazon. Your own store on another marketplace is one of those external prices.

The consequence for a split operating model is direct. When your Rakuten specialist runs a well-executed campaign discount, it has, mechanically, altered an input to your Amazon Japan listing's eligibility for the Featured Offer. It did nothing wrong; discounting during a Rakuten event is exactly what you hired it to do. But no part of its mandate covers what that does to Amazon, and your Amazon specialist finds out when the metric moves.

Two honest caveats. First, we could not verify an amazon.co.jp-published page carrying this wording — the language above is from Amazon's global seller-facing material, so treat it as Amazon's stated seller guidance rather than a confirmed Amazon Japan policy document, and read your own Pricing Health dashboard for how your listings are being assessed in practice. Second, Amazon does not publish a formula, so nobody — including any agency that tells you otherwise — can quantify the effect of a given price gap in advance. What is knowable is the direction and the governance gap, and the governance gap is the part you are deciding about.

2. Discount eligibility is a shared resource that gets consumed

This is the seam most foreign brands have never heard of, because it has no equivalent in most home markets. In Japan, whether you may display a strike-through comparison price depends on your own selling history at that price.

The Consumer Affairs Agency's guidance on price representations under the Act against Unjustifiable Premiums and Misleading Representations sets out the test. Looking back over the eight weeks before the sale begins, a past price generally counts as one "sold over a recent substantial period" when the product was sold at that price for the majority of the period it was on sale. Two exclusions sit on top of that: the price does not qualify if it ran for less than two weeks in total, or if more than two weeks have passed since the last day it was sold at that price.

Rakuten turns this into an operational gate. Its published Super SALE Search application conditions require, for a discounted item, a track record of selling at the original price for a total of at least four weeks within the eight weeks before the sale starts, plus a sale at the original price within the two weeks before it — with a shorter-history alternative of a majority of the selling period and at least two weeks. Discount listings need at least 10% off the original price; half-price placement needs at least 50%.

Read those two together and the operational reality is unusual: your reference price is depletable inventory. Every week you spend at a promotional price is a week not spent accumulating the baseline your next campaign needs, and the clock resets in a way that punishes continuous discounting. Planning a Japanese promotional year is therefore partly a budgeting exercise in baseline weeks.

Now split that planning across two agencies who each optimise their own channel's calendar. Each can run a defensible campaign schedule. Together they can consume the price history you needed for the one event that mattered most — and the failure surfaces at application time, weeks before the event, when there is nothing left to do about it. Note also that this is compliance-adjacent: the underlying rule is a consumer-protection standard, not merely a platform preference. We are describing how it shapes operations, not advising you on it; a specific display decision belongs with a Japanese lawyer or your platform's merchant support.

3. One stock pool, two promotional calendars

The third seam is the most intuitive and the most often waved away. If your channels draw on shared inventory, then every promotion is a claim on the same units. A Rakuten event and an Amazon Japan sale event scheduled a week apart are two agencies each planning a demand spike against a stock position that supports one.

Splitting stock by channel in advance sounds like the fix, and it is a partial one, but it converts a coordination problem into a forecasting problem — you now hold safety stock in two places, and the channel that undersells has stranded units the other channel needed. If your fulfilment setup already splits inventory physically, our comparison of 3PL versus cross-border fulfilment covers the structural trade-offs there.

The point for this decision is narrower: whoever allocates stock between channels is making a commercial decision about which channel is allowed to grow this quarter. That is not an operational task to be delegated to whichever vendor asks first.

Not sure who owns the seams in your current setup? Tell us which Japanese marketplaces you are live on and how the work is split today, and we will map where the cross-channel decisions are actually landing.

Talk to LAUNOVA

What Coordination Actually Costs

The invoice comparison is the easy part, and it is misleading. Two specialists usually each quote less than one multi-channel engagement, so the split can look cheaper per channel and still cost more in total. Against a market where ongoing marketplace management commonly runs in the low hundreds of thousands of yen per month per engagement — see our breakdown of Japan ecommerce pricing models for how retainer and commission structures compare — the fee difference between models is real but rarely decisive.

The cost that decides it is the one that never appears on an invoice, because you pay it in your own team's hours.

The following is LAUNOVA's own planning arithmetic, not an industry benchmark and not a quotation. We use it to size engagements; run it against your own calendar rather than treating it as a published figure.

Coordination work	One agency, 3 channels	Three specialists
Recurring review calls	1 per month	3 per month
Reporting formats to reconcile	1	3, cut differently
Contracts and renewal cycles	1	3
Parties to brief on one pricing change	1	3, sequentially
Owner of the cross-channel calendar	The agency	You
Routine internal hours/month (our estimate)	~2–3	~6–12

Two observations about that last row. It is an estimate of routine months, and event months are materially worse, because aligning a promotional calendar across parties who do not speak to each other is the single most time-expensive recurring task in a split model. And the hours are not the real constraint — seniority is. Reconciling two vendors' conflicting recommendations about pricing is not administrative work that can be handed to a coordinator. It requires someone who can overrule a vendor, which means it consumes the attention of whoever owns your Japan P&L.

This is the honest form of the argument our agency-selection guide makes more briefly: the overhead of a second agency is not the second retainer, it is that a decision which used to be made inside a vendor now has to be made inside you.

The Single-Point-of-Failure Argument Cuts Both Ways

Concentration risk is the strongest argument for splitting, and it is usually stated too loosely. Both models fail; they fail differently, and the difference is about blast radius versus frequency.

Failure mode	One agency, all channels	Specialist per channel
Key person leaves the vendor	Can affect every channel at once	Contained to one channel
Relationship breaks down	Whole Japan operation in play at once	Replace one vendor, keep the rest running
Vendor underperforms	Harder to benchmark — no internal comparison	Easier to see, you have peers to compare
Something falls between channels	Someone inside the vendor owns it	Nobody owns it unless you named someone
Migration when you do switch	One large, disruptive transition	Smaller, but you may face several over time

Two things are worth saying plainly. First, the split model's real advantage here is not resilience in the abstract, it is benchmarkability — with two vendors you can see which one is actually good, and that visibility is genuinely hard to buy any other way. Second, the concentration risk of a single agency is largely a contracting problem rather than a structural one: account documentation, admin-level access held by you, a defined handover obligation and a notice period sized to a real migration convert "one throat to choke" from a liability into what it is supposed to be. If you are weighing that transition cost, our guide to switching a Japan EC agency covers what a handover involves in practice.

Where a Specialist Genuinely Wins

We sell multi-channel operations, so read this section knowing which answer is commercially convenient for us. There are cases where splitting is right, and they are specific rather than general.

One channel is disproportionately large. When a single marketplace carries most of your Japan revenue, its advertising and merchandising alone can justify a dedicated team, and the seam cost is worth paying to get depth where the money is.
A channel requires a genuinely different discipline. The visibility mechanics are not variations on a theme — Rakuten's RPP is a click auction, while Yahoo! Shopping's main visibility lever is billed as a percentage of the sale, which makes setting it a margin decision rather than a media-buying one. Our comparison of Japan EC advertising sets out how far apart these actually are.
You already have a Japan-side owner. If someone on your payroll owns pricing and the promotional calendar and can overrule a vendor, the seams have a home and the main argument against splitting weakens considerably.
You are testing a new channel. A short specialist engagement to prove a channel is worth having is a cheaper experiment than expanding an existing contract, and it is reversible.
You need a benchmark. If you suspect your incumbent is coasting and cannot tell, putting one channel with a specialist buys you a comparison. That is a legitimate reason to split even if you consolidate again later.

A Framework for Deciding

Rather than a revenue threshold — which we would be inventing — the useful test is a sequence of questions in order. The first "no" is your answer.

Is there a named person, on your side, who owns pricing and the promotional calendar across all channels? If no, do not split. The split model has an unpriced dependency on this role, and running it vacant is how the seams above turn into losses.
Is any single channel large or distinct enough that a dedicated team would change its trajectory? If no, splitting buys depth you have no use for and pays for it in coordination.
Can you absorb the coordination load at a senior level, every month, including event months? If no, you will get the split model's costs and, in practice, the generalist model's depth — the worst combination.
Would a specialist genuinely outperform, or does the incumbent merely need a better mandate? Underperformance on a channel is often a scoping failure rather than a capability failure. It is cheaper to test a rewritten mandate than to run a migration.
If you split, what is written down about the seams? If nothing, you have not split responsibilities — you have dispersed them.

A brand that answers yes to the first three is usually ready to split at least one channel. A brand that answers no to the first is not ready regardless of size, and the honest recommendation is to consolidate and fix the mandate.

What to Put in the Contract, Either Way

Most of the damage described above is preventable by writing down a small number of things that are almost never written down. This list applies to both models — in a single-agency setup it defines what you have delegated, and in a split setup it defines the boundary that would otherwise be unowned.

Pricing authority. State who may change a listed price, within what band, and with what notice to the other channels. If two vendors each hold unbounded pricing authority on their own channel, you have not defined pricing authority at all.
Promotional calendar sign-off. A single calendar covering every channel, approved by you, with a rule for what happens when two events collide. Include the baseline-price weeks each planned campaign requires, so eligibility is a scheduling input rather than a discovery at application time.
Inventory allocation. Who decides the split between channels, on what cadence, and who is consulted before a campaign that assumes an allocation.
Reporting comparability. If you run multiple vendors, specify the metric definitions and the period cut. Otherwise you will spend the coordination hours above reconciling numbers rather than acting on them.
Access and exit. Admin-level platform access retained by you, an account-documentation obligation, and a notice period long enough for a real handover.

If you want the vendor-evaluation side of this — the questions to ask, and what a weak answer sounds like — our comparison of Japan EC agencies covers the market landscape and includes the related but separate question of whether an agency covers owned storefronts alongside marketplaces. That is a scope question about what a single agency sells; this article is about how many agencies you buy from.

What LAUNOVA Does — and Where We Are Not Neutral

We should be direct about the conflict. LAUNOVA runs multi-channel marketplace operations for overseas brands, so an article concluding that one partner across all channels reduces coordination risk is an article arguing for the thing we sell. Treat the fourth question in the framework above as the check on that: if your incumbent's problem is a bad mandate rather than bad capability, rewriting the mandate is cheaper than hiring anyone, including us.

What we actually run is the Japanese-language operating layer across Rakuten, Amazon Japan and Yahoo! Shopping as one coordinated operation — with the seams above held explicitly rather than by accident. What we do not do is set your margin, take pricing authority you have not delegated in writing, or claim that a single vendor removes concentration risk; it relocates it, and the mitigation is contractual. We are also not a law firm: the price-representation rules described here shape how a promotional calendar has to be planned, but a decision about a specific display belongs with a Japanese lawyer.

If you are weighing this now, the most useful things to send are which channels are live, how the work is split today, and which handoffs have failed in the last six months. We will tell you where the seams are landing — and if the answer is that you need a specialist on one channel and us on the rest, we will say that. Get in touch, or read how remote Japan operations support is scoped if you want the delivery model first.

How to Choose a Japan EC Agency

The evaluation criteria for judging any single agency — platform coverage, content quality, compliance and reporting.

Japan EC Agency Comparison

Who is actually in the market, and the separate question of marketplace-only versus full-service scope.

Switching Your Japan EC Agency

What a handover involves — the migration cost that sits behind every consolidate-or-split decision.

Sources

• Comparison-price rules — a past price generally qualifies as one sold over a recent substantial period when, within the eight weeks before the sale begins, the product was sold at that price for the majority of the period it was on sale; it does not qualify if that price ran for less than two weeks in total, or if more than two weeks have elapsed since it was last sold at that price — Consumer Affairs Agency, 不当な価格表示についての景品表示法上の考え方 (guidance on price representations under the Act against Unjustifiable Premiums and Misleading Representations), section on 過去の販売価格を比較対照価格とする二重価格表示. Primary source.
• Rakuten Super SALE Search application conditions — original-price sales record of at least four weeks total within the preceding eight weeks, or a majority of the selling period and at least two weeks where the item has been on sale for less than eight weeks; plus an original-price sale within the two weeks before the sale starts; discount placement requires at least 10% off and half-price placement at least 50% off — Rakuten Ichiba official merchant help page (ichiba-smp.faq.rakuten.net). Primary source.
• Featured Offer and external price comparison — Featured Offer prices are commonly at or below the lowest priced alternatives, and Pricing Health guidance directs sellers to price at or below the competitive external price — Amazon seller-facing material (sell.amazon.com). Amazon's own guidance, but not verified against an amazon.co.jp-published page; treated in this article as stated global seller guidance rather than a confirmed Amazon Japan policy statement. Amazon publishes no formula for the effect of a given price gap, and none is asserted here.
• Super SALE application window (roughly three weeks to one week before the event) — Japanese seller and agency guides, second-hand; confirm the current window in your own RMS notices, which is where the authoritative dates appear.
• Coordination-hours table and the two-to-four hours per additional vendor figure are LAUNOVA's own planning arithmetic, not an industry benchmark, survey or quotation. They are illustrative sizing assumptions; run them against your own calendar.
• Ongoing marketplace management fee ranges, Rakuten RPP click-auction mechanics and Yahoo! Shopping's percentage-of-sale promotion option are carried over from our existing published guides rather than re-sourced here; see the linked articles for their underlying sources.

One Agency for Every Japan Marketplace, or a Specialist per Platform?