Content density chart of this comparison: words in each answer against websites cited
← Comparison · Top 5 crowdfunding channels in Japan for overseas food providers: 6–12 month total project costs AI answer · one of 9

Opus 5.5 at Medium, asked 8 Oct 2026

One answer from the comparison, shown in full as the model produced it on 8 October 2026. Nothing is edited. Its 13 cited links have not been opened, so this page does not say whether the pages back what it wrote.

AI-generated · unverified · asked 8 Oct 2026 · links not checked

The answer · as produced

I'll check your food-product context, then research the main Japanese crowdfunding platforms' fees, ad options, and agency costs to build a 6 to 12 month total cost comparison.

I've ranked the top 5 below by how well each fits an overseas food brand. The main finding is that the platform commission is only about a fifth to a quarter of what a 6 to 12 month crowdfunding program costs in Japan. Ads, fulfillment, the import setup and average order value change your total cost more than the gap between platforms does.

Top 5 channels for an overseas food brand

1. Makuake: the default for a "new product debut" Makuake is the biggest market by money raised. For January to May 2026, Makuake raised about ¥7.87B, a 44.2% share of the market, while CAMPFIRE led on project count with 65.5%. (craco via PR TIMES, 2026-06-15)

Its official fee is 20% (tax excluded) of the backed amount for Makuake projects, and an optional paid ad support package exists for projects that meet internal criteria. (Makuake legal notice, accessed 2026-10-08)

For overseas makers, Makuake runs a Global Official Partner Program. Makuake introduces a partner that acts as the project's operating agent, handling page creation, pre-launch promotion and backer support through delivery. (ECのミカタ, date not shown, accessed 2026-10-08)

One caveat: Makuake requires projects to state when a product is an import-agency or overseas OEM item. Your story needs to be about what is new for Japan, not just "imported." (Impress Watch, accessed 2026-10-08)

2. CAMPFIRE: cheaper, high volume, less hand-holding CAMPFIRE's standard fee has been 17% (tax excluded). (ASCII, accessed 2026-10-08) Of that, the payment portion is 5% of the backed amount, and a separate per-backer system fee is charged. (ネットショップ担当者フォーラム, 2026-07-30)

CAMPFIRE has a large audience, about 394,000 backers in Jan to May 2026, a 40.7% share. The weakness is that its project count is huge, so it is easy to get buried without your own traffic. (craco via PR TIMES, 2026-06-15)

3. Kibidango ("Kibitan" program): the only model that takes over import and logistics This is structurally different from the others. Kibidango becomes the seller and manages import and logistics, and also handles backer support. The tradeoff is that its track record is almost entirely gadgets and goods. A food brand would be an exception for them, so confirm fit before planning around it. (Kibidango Kibitan page, stats as of 2026-10-08)

Kibidango has historically advertised a 10% fee, but that figure comes from a 2016 release. Under the Kibitan model, your real cost is the margin you give up in a negotiated split, and that split is not public. (value-press, 2016)

4. GREEN FUNDING: only for premium, high-ticket food-adjacent items It had only 107 projects in Jan to May 2026, but averaged about ¥13.69M each, more than 7 times the market average of about ¥1.82M. That average reflects a gadget-heavy audience. It fits only if your item is premium kitchen or beverage gear rather than consumable food. (craco via PR TIMES, 2026-06-15)

A comparison site lists its Standard plan fee at 18.2%. I could not confirm this on an official page, so treat it as indicative. (gooddo, date not shown, accessed 2026-10-08)

5. READYFOR: only if there is a social or regional story READYFOR is strong in social causes. "READYFOR × Social" raised about ¥1.71B in Jan to May 2026, the fourth-largest platform-category combination. (craco via PR TIMES, 2026-06-15)

It is a weak fit for a commercial food launch unless there is a producer-community or sustainability angle. I did not verify its current fee.

Kickstarter is not in the top 5. Its fees are low, 5% plus roughly 3 to 5% payment processing, but its backers are not Japanese consumers. (Kickstarter Help Center, accessed 2026-10-08)

The cost that applies to every platform: import compliance

Food is harder than gadgets in one way that holds on any platform. Food imported for sale requires an import notification by the importer under Article 27 of the Food Sanitation Act, and food without that notification cannot be sold. (MHLW, accessed 2026-10-08)

The importer also carries the same legal responsibility as a domestic manufacturer or seller. (JETRO, accessed 2026-10-08)

So you need a Japanese importer of record before any campaign. Kibidango is the only channel above that takes on that role itself.

Total cost model for 12 months

The agency rates come from a 2026 listing for a crowdfunding consultancy: production and planning support of ¥300k to ¥1M, a success fee of 10% of the amount raised, an average of ¥1M to ¥1.5M per project, and about 6 months per project. (Batonz listing, updated 2026-05-29)

Everything else in the table is my estimate, not sourced data. The scenario is two campaign waves of ¥3M each, so ¥6M total, with an average order of ¥5,000, which works out to about 1,200 orders.

Cost lineMakuakeCAMPFIREBasis
Platform commission (incl. 10% tax)¥1.32M (22%)¥1.12M (18.7%)Official rates above
Agency or partner (production plus 10% success fee)~¥1.25M~¥1.25MBatonz 2026 listing
Ads (assumed 10 to 25% of GMV)¥0.6M to ¥1.5M¥0.6M to ¥1.5MMy assumption
Fulfillment (assumed ~¥1,200 per order, shelf-stable)~¥1.44M~¥1.44MMy assumption; chilled costs more
Importer and label setup~¥0.5M~¥0.5MMy assumption
Total before product cost (COGS)~¥5.1M to ¥6.0M~¥4.9M to ¥5.8M

Four things stand out from this:

  • Choosing a platform saves about ¥200k over 12 months. That is around 3 percentage points of GMV. Fulfillment and ads each cost more than that difference.
  • Average order value is the biggest lever. At ¥5,000 per order, fulfillment alone eats about 24% of GMV. A ¥10,000 bundle cuts that roughly in half.
  • At ¥6M GMV, channel costs consume most of the revenue before product cost. At this scale, crowdfunding works as paid market validation, not as profit. Budget it that way.
  • Kibidango shifts most of the import, fulfillment and support cost onto them, in exchange for a margin share. For a limited budget, that makes it worth an inquiry even though its catalog skews to gadgets.

Using a campaign as evidence for Japanese buyers

Since the goal is to adapt the product line for Japanese buyers, a better use of the money than "testing" is to run one Japan-specific SKU on Makuake, ideally a bundle with a high average order. The campaign then gives you backer counts, sell-through by variant and review language. That data becomes your pitch to retail buyers in months 6 to 12, which is the part a pure test never produces.

Sources:

End of the answer. AI-generated, unverified. Asked 8 Oct 2026.

Counted · websites it cites

13 websites, 13 links.

The same count the comparison's density chart uses. A website is a distinct hostname. None of these pages has been opened and compared with its passage.

WebsiteLinks
ascii.jp1
batonz.jp1
ecnomikata.com1
gooddo.jp1
help.kickstarter.com1
jetro.go.jp1
kibidango.com1
makuake.com1
mhlw.go.jp1
netshop.impress.co.jp1
prtimes.jp1
value-press.com1
watch.impress.co.jp1

A link counts once however often the text repeats it.

The same question · the other 8 answers

Read another answer.

ModelSettingAskedWordsWebsites
Sonnet 5.5AnthropicMedium8 Oct 20261,15617
GPT-6 LunaOpenAIMax8 Oct 20261,24612
GPT-6.1 SolOpenAIMax8 Oct 20261,74119
Gemini 3.6GoogleThinking8 Oct 20264,26025
Grok 4.7xAIFast8 Oct 20261,30411
Qwen3.8-MaxAlibabaThinking8 Oct 20261,134–
Qwen3.8-Omni-FlashAlibabaThinking8 Oct 20261,661–
Haiku 5.5AnthropicExtra8 Oct 202684114
The research plan
$50 first month, then $200 a month

Want the sources checked for your product?

This answer is unchecked. In the research plan, we open the sources, check the dates, and correct what is wrong before the brief reaches you.