The Japanese Customer Reference Interview Process, Start to Finish

What "how it's made" looks like when you don't hide it

A customer reference is not a writing task with an interview attached. It's an interview with a writing task attached — and four other steps wrapped around both. Showing you all six is the point of this page.

Here is the full pipeline, top to bottom:

1. Source the reference and secure 許諾 (permission) — find the right customer and get them to a clear yes.
2. Run the native-language interview — the hour where the story actually gets made.
3. Write for a Japanese B2B reader — not translate, write, in the register a Japanese buyer trusts.
4. Produce the English report back to HQ — so a non-Japanese-speaking manager can see the quality.
5. Carry it through approval — the 稟議 clock you don't control.
6. Deliver, with a fallback — the named version, or a credible anonymous one if approval stalls.

Most engagements quietly skip steps 2, 4, and 6 — they treat the interview as a phone call, drop the English report, and have no plan for when a customer says "not under our name." Those are exactly the steps that decide whether the finished story persuades a cautious Japanese buyer and reassures your head office. So those are the ones this page spends the most time on.

Step 1 — Sourcing the reference and securing 許諾

Before anyone interviews anyone, two things have to happen: pick the right customer, and get a real yes.

Picking the right customer is a marketing decision, not an admin one. The best reference is rarely your happiest customer — it's the one whose before-state most resembles your next prospect's. A mid-size manufacturer who solved a boring, expensive operational problem persuades more buyers than a flashy logo who can only say "great product." Sourcing means choosing the story that does the most selling, then confirming the customer is articulate enough to tell it.

Securing 許諾 (permission) is where overseas teams underestimate Japan badly. In most Western markets, a customer's marketing contact says "sure, happy to" and you're done. In Japan, that verbal yes is the start of approval, not the end. The named individual may be willing, but their company's name, logo, and quotes usually have to clear an internal review — and the person you're talking to often can't grant that alone.

So this step produces two things:

• A scoped yes — agreement on exactly what will be public (company name? individual's name? logo? specific numbers?), in writing, before the interview, so nobody is surprised later.
• A realistic approval map — who internally has to sign off, so the timeline (Step 5) isn't a shock.

Getting this clear up front is what prevents the most demoralizing outcome in Japanese case-study work: a great interview, a great draft, and then a customer who quietly withdraws their name because nobody warned them what "published" actually meant. We go deep on why that yes is so hard to get — and how to design around it — in Why Japanese Customers Won't Agree to a Case Study.

Step 2 — The native-language interview (what good questions sound like)

This is the step everything else depends on, and the step that cannot be outsourced to translation or AI.

A customer reference is only worth publishing if it contains specifics: the real problem before the product, the hesitation during evaluation, the concrete change afterward. Japanese B2B buyers are unusually alert to stories that sound like marketing — vague praise reads as fabricated, and fabricated-sounding proof is worse than no proof at all. The interview is where you either get those specifics or you don't.

Getting them requires a native interviewer who can do three things a translator can't:

• Build enough comfort that the customer says the true thing, not the safe thing. A Japanese executive's first answer is almost always the polished, company-line version. The real story comes out on the third follow-up, after they trust you enough to admit what was actually hard.
• Follow a hesitation. When a customer pauses or softens an answer, that's usually where the interesting truth is. A native interviewer hears the hedge; a script doesn't.
• Read what isn't said. In Japanese business conversation, the most important thing is often implied rather than stated. Surfacing it tactfully — without putting words in the customer's mouth — is a craft.

What do good questions sound like? Not "How satisfied were you?" (which only ever gets you "very satisfied"). They sound like:

• 「導入を決める前、いちばん不安だったことは何でしたか?」 — What worried you most before deciding? (Surfaces the real objection your next prospect also has.)
• 「他社製品とも比較されたと思いますが、最後の決め手は?」 — You probably compared other products; what was the final deciding factor? (Gets the honest, often unglamorous reason.)
• 「導入直後、現場の反応はどうでしたか?」 — How did the team on the ground react right after rollout? (Reveals the real adoption story, not the executive summary.)

Each question is built to pull out a specific, defensible detail — a before-state, a comparison, a concrete result — rather than a compliment. That hour is where the value of the entire reference is created. The writing in Step 3 documents a good interview or papers over a bad one; it cannot manufacture material that the interview failed to capture.

Step 3 — Writing for a Japanese B2B reader, not translating

With a good interview captured, the next step is writing — and "writing" is the precise word. This is not transcription, and it is emphatically not translation.

A Japanese B2B reader trusts a customer story for reasons that don't transfer from English. They want to see the before-state stated plainly. They want a number they can map onto their own situation. They want a quote that sounds like a real person making a slightly unglamorous, professional decision — not a marketing department's idea of enthusiasm. And they are immediately suspicious of copy that reads like it was written to impress rather than inform.

This is why "we'll have our translator write it up from the notes" produces a flat, generic story at a fraction of the cost and roughly none of the persuasion. Translation moves words across a language. Writing-for-the-reader moves trust across a culture — it decides what to lead with, what to cut, which quote to feature, and how to frame the result so a cautious buyer believes it.

Here is what that difference looks like in practice. This is an illustrative composite — not a real client, with names removed but the persuasive specifics kept — showing the kind of flat draft a notes-and-translation workflow produces, and the version that actually persuades a Japanese buyer.

Step 4 — The English report back to HQ

Here is the step almost no provider includes, and the one a first marketer needs most.

You ran a great interview. You published a sharp 導入事例. Now your head office asks, in English: Is this good? What does it actually say? Did the customer endorse us, or just agree to be named? If you can't answer that crisply, the reference you worked weeks to land becomes a source of anxiety instead of confidence — the work is invisible to the people who renew your budget.

The fix is a bilingual handoff built into production itself. Alongside the Japanese story, a short English report that:

• summarizes what the customer actually said — the before-state, the deciding factor, the result, in plain English;
• flags the strongest quotes and explains why they land for a Japanese buyer;
• explains the framing choices — why this story leads where it does, and where it deliberately departs from how a US case study would read.

This is not a back-translation. Back-translations are clumsy and miss intent; they tell your head office what the Japanese words are, not whether the story is any good. A practitioner's English report tells a non-Japanese-speaking manager whether the work is strong — so they can see the quality for themselves and defend the spend.

Most freelance translators can't produce this, because it requires understanding the marketing intent, not just the language. Most agencies charge it as a separate engagement, if they offer it at all. Native interview in, clear English report out, as one workflow — that combination is the rarest part of this pipeline and the reason it's worth more than a translate-the-notes alternative. We go deeper on the reporting side in Reporting to an English-Speaking Head Office.

Timeline, deliverables, and what you receive

So how long does the whole pipeline take, and what shows up at the end?

Two clocks run on a Japanese reference, and confusing them is how timelines blow up.

The production clock is the part the provider controls — sourcing, interview, writing, English report. End to end, that's roughly two to four weeks for a single piece, most of it scheduling the interview and the revision pass, not the writing itself.

The approval clock is the part you don't control (Step 5, below), and it can run three to six weeks on top — sometimes concurrently, sometimes not.

The realistic deliverables for one credible reference:

The honest version of "what you receive" is: one published story your buyers trust, one English report your head office trusts, and a documented process you can repeat. The piece is the visible output. The repeatable pipeline is the asset.

Step 5 — Carrying it through approval (the 稟議 clock)

Two things surprise overseas teams about Japanese approval, and both belong in the process, not the fine print.

First, the approval clock is separate from the editorial clock. Editing is normal: draft, your feedback, one or two revision passes, sign-off — predictable. But before a Japanese customer's name, logo, or quote can go public, it usually has to clear an internal 稟議 review: legal, the named executive, sometimes corporate communications. This is not foot-dragging; it's how careful Japanese companies protect their brand. It means the gap between "draft you approved" and "live on your site" can be three to six weeks.

Second, the answer that comes back isn't always yes. Sometimes it's "anonymized only," or "not yet," or "the name is fine but remove that number." A process that hasn't planned for this stalls the whole content calendar on one customer's legal department.

Two design rules keep this from breaking your program:

1. Build in time, not just cost. Never promise HQ a two-week named case study end to end. Plan the pipeline so one slow approval doesn't freeze everything behind it.
2. Always have an anonymous fallback ready (Step 6) — so a stalled approval still ships as a usable asset instead of becoming a dead loss.

Step 6 — Deliver, with a fallback when 許諾 stalls

The pipeline ends in delivery — but a mature process delivers a usable asset even when the named version doesn't clear approval.

If a Japanese customer's name gets stuck in 稟議 (or comes back as "anonymized only"), a well-built anonymous version still persuades. Japanese buyers routinely read named and unnamed stories with the same scrutiny — what they trust is the specificity, not the logo. The mistake is letting "we lost the name" turn into "we lost the story," when the fix is editorial: keep the industry, the headcount, the before-state, the numbers, and the real quote; remove only what identifies.

Done well, the fallback isn't a downgrade — it's a parallel deliverable the process produces on purpose. We rebuild a flat anonymous story into a convincing one, step by step, in Why Japanese Customers Won't Agree to a Case Study (And How to Get One Anyway).

That's the full pipeline: source and secure 許諾, interview in Japanese, write for the reader, report to HQ in English, carry it through approval, and deliver — with a fallback that keeps the work from being wasted. Six steps, none hidden.

Per-piece vs an ongoing reference pipeline

Now the budget question: do you buy one reference, or build a pipeline?

One reference, once, to test whether customer stories move your Japanese buyers, is a legitimate way to start. Treat it as a learning spend — you'll find out fast whether the format earns its cost in your market.

But in Japanese B2B, the customer reference is often the single most-requested asset before a buyer will move — which means one is rarely enough. And per-piece economics fight a pipeline: every story re-pays the setup cost of choosing the customer, securing 許諾, briefing an interviewer, and producing the English report from scratch. The interview muscle never stays warm; the approval relationships never compound.

This is why we publish flat monthly pricing rather than a per-reference quote:

Month-to-month. A senior practitioner working with you directly, no account-manager layer. SOW, NDA, and company profile prepared same-day for your internal approval process.

For comparison, here's what the market quotes for this work — and what it usually leaves out:

The decision rule is simple. One story to test the format? A single defined project. Two or three a year on a predictable cadence? Growth. Customer references as a standing pipeline — your most-requested sales asset, produced continuously and reported to HQ in English? That's an Embedded function, and at that volume per-piece pricing is both slower and more expensive, because every story re-pays the setup the first time taught you to avoid.

You can see the full economics, with the per-piece math laid out, in How Much Does a Japanese B2B Case Study Cost?

Where to go next

• How Much Does a Japanese B2B Case Study Cost? — the full economics, with the per-piece math laid out
• Why Japanese Customers Won't Agree to a Case Study (And How to Get One Anyway) — the 許諾 problem, and the anonymous fallback done right
• The First Marketer in Japan: An Operating & Budget Playbook — the pillar this pipeline plugs into

See it in practice: Japan Content Teardown →