Frequently asked questions
Price, timing, targeting, data handling and refunds — 24 questions about AI-moderated interviews with a Japan-resident, Japanese-first-language panel. If yours is not here, ask us; we answer in English.
Pricing and payment
What does $49 actually pay for?
Recruiting a participant who matches your criteria, screening them, running the interview in Japanese, translating it, writing the English session summary, contributing that session to the cross-study report, and handling the participant's payment inside Japan. It is charged once per completed session.
See the full price list →What does a whole study cost?
The platform fee plus the incentive you choose, multiplied by the number of sessions. 10 30-minute sessions with a $21 incentive is 10 × ($49 + $21) = $700. 5 sessions with a $14 incentive — the smallest study possible — is $315. The exact total is shown before you pay.
See the full price list →Why is the incentive shown in dollars when it is paid in yen?
Because you are charged in dollars and the participant is paid in yen. We convert the yen amount at the rate in effect when you order and add a small buffer, so that a currency movement between your payment and the participant's payout does not reduce what they receive. The participant gets the full yen amount; we take no share of the incentive.
Is there a subscription, a seat fee or a minimum commitment?
No. There is no monthly fee, no per-seat charge, no annual contract and no minimum spend beyond the 5-session minimum for a single study. You pay per study, when you run one.
How do I pay?
By card, through Stripe, in US dollars. We do not currently support bank transfer, purchase orders, invoicing in arrears or payment in other currencies. If your company cannot pay by card, contact us before you start.
Contact us →Will I get an invoice for my accounting system?
Yes. Stripe issues a receipt with UZRA Inc.'s registered details at the moment of payment, and it is available from your dashboard afterwards. Prices are tax-inclusive — the amount you see is the amount charged, with no tax added on top.
Can I get a refund?
Yes, for anything you did not use. You can stop recruiting at any point and the platform fee and the incentive for every unused session are refunded to the card you paid with. A session that a participant abandons, or that fails our quality review, does not consume a slot — recruiting continues instead. If a session that was delivered to you turns out to be unusable, tell us and we will look at it.
Read the terms →Running a study
How do I start?
Create an account, write your research objective in English, and we draft a Japanese interview guide from it. You review the guide in English, edit anything, set your audience, and see the price. Nothing is charged until you confirm.
How long does a session take?
Every session runs about 30 minutes. The interview is a text chat and typically covers 8 to 24 exchanges, ending when the guide has been worked through rather than at a fixed clock time.
How long does the whole study take?
Recruiting starts the moment you pay, and sessions arrive one at a time as participants complete them — you do not wait for the study to finish before reading anything. How fast that goes depends entirely on how many panel members match your criteria, which is why the audience screen shows you an estimate before you pay. We do not promise a completion date, and we do not ask you to carry that risk: anything unfilled is refunded.
How many participants do I need?
In a single, coherent audience, most teams find that themes start repeating around the eighth to tenth interview, and ten is a reasonable default. 5 is the minimum and is enough to test a strong, specific hypothesis. If you are comparing two audiences — say lapsed users against active ones — treat each as its own study rather than splitting ten sessions between them.
Read the recruiting guide →Can I run the same study again later, or talk to the same people again?
You can run the same guide again with new participants at any time. You cannot reach the same people again: participants are identified to you only by a study-scoped ID, and re-contacting them is not permitted under our terms. Longitudinal and diary studies are not something we support.
Can I watch the interview happen?
No. Sessions are asynchronous — a participant starts one when it suits them, without you or us present. You see the full transcript afterwards.
Platforms that do run live sessions →The panel
Who is in the panel?
People who live in Japan and use Japanese as their first language. They joined through the Japanese-language side of torima.in, which has operated as a research marketplace since 2020, and they sign up, apply, take part and get paid entirely in Japanese. This is the whole point of the service: a panel that is reachable only in Japanese does not fill up with the English-speaking minority.
What can I target on?
Registered profile attributes — gender, age band, region, occupation and household income — with a live count of matching panel members as you narrow. On top of that you can write screening criteria in free text for anything a profile field cannot answer. Those are declared by the participant when they apply and confirmed by the interviewer at the start of the session.
Why is my screening criterion not reflected in the panel-size estimate?
Because we have never asked the panel that question, so we genuinely do not know the answer. Showing you an estimate that pretended to include it would be a made-up number. The estimate covers the profile attributes only, and narrows honestly.
Can I reach B2B or specialist professional audiences?
Sometimes, and you will see before you pay whether we can. This is a consumer panel; narrow professional audiences — enterprise decision-makers, licensed specialists, particular industry roles — are the ones we most often cannot fill. If the estimate is low, do not order and hope.
Where a B2B panel is the better answer →How do I know the participants are real, and answering seriously?
Every completed session is scored automatically against the guide: whether the participant engaged with the topics, whether there is enough substance in what they wrote, and whether their screening answers held up under questioning. Sessions that clear the bar are delivered and the participant is paid. Sessions that do not are held and read by a person on our team before any decision, and can be rejected — in which case the slot is released and recruiting continues. Participants also carry a rating history from the wider torima.in marketplace.
The AI interviewer
What is an "AI-moderated interview"?
A conversation, in Japanese, between the participant and a language model working from the guide you approved. It asks an opening question, listens, follows up on what the participant actually said rather than reading the next scripted item, and works through your topics. It is not a survey with an open-text box.
What can the AI interviewer not do?
Be honest with yourself about these before you order. It cannot see the participant, so tone, hesitation and body language are lost. It cannot react to a screen share or watch someone use a prototype. It will not improvise a whole new line of enquiry that is nowhere in the guide — it probes within your topics, it does not replace a researcher's judgment about what the study should have been about. And a text chat selects for people who are comfortable writing, which is a real sampling effect. For high-stakes generative work at the start of a project, a live moderator is still better. This is a way to get many more conversations, sooner and cheaper, than that.
When to use a live moderator instead →Who writes the questions?
We draft them from your objective; you approve them. You can rewrite any question, add your own, reorder them, or regenerate the whole guide. You review it in English, and it runs in Japanese.
How good is the translation?
Summaries, quotes and transcripts are machine-translated and reviewed for meaning before delivery. The original Japanese is kept and available for every transcript and every quote, so if a phrase matters to your decision you — or a Japanese-speaking colleague — can check it against the source. We would rather you verify than take our word for it.
Your data and participants' data
What personal data do I receive about participants?
None. You get a study-scoped participant ID and the attributes you filtered on — gender band, age band, region, occupation, household income. Names, contact details, exact ages and free-text identifiers are stripped before anything reaches your dashboard, and our terms prohibit you from attempting to identify, contact or re-target a participant, or from using the data to build a profile of an individual.
Read the privacy policy →How does this sit with Japan's APPI and with GDPR?
Participants are recruited, interviewed and paid entirely inside Japan, and their personal data stays with us in Japan. Before anything reaches your dashboard, responses are de-identified: you see a study-scoped participant ID and the coarse attributes you filtered on, never names or contact details. Participants consent in Japanese to their anonymised, translated responses being shared with the commissioning company. Japan holds an EU adequacy decision. We do not claim this removes every obligation on your side — if your privacy team wants to review the setup before you run a study, contact us and we will walk them through it.
Contact us →Still deciding?
The guide to recruiting Japanese research participants covers all six routes, including the five that are not us. Or start a study and see the exact price before you pay anything.
Questions first? Contact us — we answer in English.