What are Japanese players actually saying about your game?
Steam shows you a Japanese score. It does not show you why — which topic dominates Japanese reviews this month, what Japanese creators and media picked up, or what to do next.
Japan Steam Radar is a monthly, English-language report for small Steam teams: your game's Japanese reviews, classified and quoted — plus Japanese media/creator signals and at most three concrete next moves.
Currently in research phase. No subscription is on sale yet. Join the waitlist below (prices are shown before you sign up), or apply for a one-off paid pilot report.
Japanese reviews are a small share of your total (for PEAK: 1,113 of 313,072 reviews, 0.356% on 19 Nov 2026, same-day same-method snapshot), so a Japan-specific problem can dominate Japanese discussion and still be invisible in your global theme list.
Japanese reviewers rated PEAK 88.4% positive vs 94.6% across all languages — a 6.2-point gap that Steam now shows you for free. What it does not show is why the gap exists, whether it is moving, or what to do about it.
Machine translation tells you what a review says. It does not tell you whether the topic is Japan-specific or global, whether it moved this month, or whether a Japanese holiday, creator event, or media piece is the real context.
The expensive alternatives are full localisation / publishing retainers (typically $2,000+/month, quotes on request). The free alternatives (SteamDB, Steam's language-specific scores, DIY AI summaries) give you numbers and raw text — not a classified, quoted, Japan-context reading.
What you get each month (per title)
Japanese review classification — Your Japanese Steam reviews, classified by topic (e.g. difficulty/balance, bugs, performance, session/matchmaking, language/UI), with counts and short quoted examples in Japanese + English translation. Topics are separated into Japan-specific vs shared with global reviews — because the fix differs.
Japanese media & creator signals — Japanese-language media coverage and notable Japanese creator/streamer moments for your title that month, with sources. Detection via YouTube and Japanese media; see Scope for what we do not cover.
Category watch — Which comparable titles in your category added Japanese, ran Japanese creator events, or changed Japanese store copy that month.
Next moves (max 3) — Concrete, costed suggestions tied to the findings above (store copy, patch-note localisation, calendar timing, known-issues posts). No generic advice unconnected to your data.
Format: one English report per title per month, fixed format. Delivered as a document. No dashboard, no meetings, no Slack channel.
Scope — what this is NOT (read before you sign up)
Not a localisation service. We do not translate your game or store page. We tell you what Japanese players are saying and what it means.
Not comprehensive X/Twitter monitoring. X does not offer a compliant, economically viable collection route for this price point, so we do not promise it. Japanese X reaction is spot-checked manually only when a specific event makes it relevant, and is labelled as a spot check.
Not a sales guarantee and not consulting. No custom research, no bespoke cuts of the data, no strategy calls. If you need a custom engagement, you need a different provider — we will say so rather than sell you this.
Not raw data resale. You receive classification, quotes (short excerpts), and interpretation — not a dump of scraped data.
Sample — what a report actually looks like
We publish full samples, not mock-ups. The PEAK full sample (v0.2, 19 Nov 2026) passed an internal red-team check on six conditions: direct classification with quotes, a dated 90-day forward calendar, a money-side estimate with method, at least two primary Japanese media sources, next moves with execution costs, and an honest free-alternatives table.
PEAK (Landfall + Aggro Crab) — full sample, headline findings
Full census, not a sample of reviews: all 1,113 Japanese reviews (17 Jun 2025 – 17 Nov 2026) retrieved via Steam's official API and classified. Difficulty mentioned explicitly in 158/1,113 (14.2%). The new "Deep Forest" biome in 27/1,113 (2.4%) overall — but 19/424 (4.5%) in the recent window (Nov 2025 – Nov 2026) and 11 of 52 recent negative reviews. Bugs in 38/1,113 (3.4%, falling from 4.4% early to 1.9% recent). Performance (narrow) in 22/1,113 (2.0%, falling from 3.0% early to 0.2% recent). "Japanese language" mentioned in 10/1,113 (0.9%) — all in the early window, zero in the recent window.
What that means: the Japanese complaint centre moved — from early bugs / performance / language defaults to one specific biome's balance. A one-off look, or a free score, does not show that movement. A monthly classification does.
Scale, with method stated: if Japanese buyers review at the global average rate, PEAK's 0.356% Japanese review share implies roughly 22,000 Japanese-attributable units (~¥19M gross at ¥880) on a 6.19M-unit estimate, or roughly 50,000 units (~¥44M) on a 14.1M-unit median estimate. Review rates differ by language, so treat this as a scale range, not measured sales.
Category context (same day, same method): Japanese review share — PEAK 0.356%, R.E.P.O. 0.514%, Lethal Company 0.149%, Content Warning 0.232%.
Forward look: the sample lists four dated Japanese windows in the next 90 days (23 Nov 2026 Labor Thanksgiving Day; 17 Dec 2026 – 4 Jan 2027 Steam Winter Sale; 1–3 Jan + 11 Jan 2027 New Year / Coming of Age Day; 11 Feb 2027 National Foundation Day) and ties each next move to a window and an execution cost in person-hours.
Full census: all 1,804 Japanese reviews classified. Japanese share 0.514% of 350,812 total reviews; Japanese positive rate 92.6% vs 96.1% across all languages.
The localisation story, corrected by the data: official Japanese arrived only on 29 Oct 2026 (v0.5.0), ~20 months after launch — yet 1,764 Japanese reviews already existed before localisation (played in English or via a community mod). Every explicit "Japanese language" mention (60) and mod mention (61) sits in the pre-localisation window; after localisation, zero. The demand was visible, in Japanese, for ~20 months.
Same pattern as PEAK: the complaint centre moved over time, and the newest map/biome content (Sunstone) accounts for 5 of the 11 post-localisation negative reviews — the R.E.P.O. counterpart to PEAK's Deep Forest.
Honest note: a full sample proves the report can be produced. It does not prove anyone will pay $49/month for it. That is what the waitlist and the one-off pilot below are for — and we count only sign-ups made after seeing the prices.
How Radar differs from free / cheaper tools
Question
Free / cheaper route
Japan Steam Radar
"What is my Japanese score?"
Steam language-specific scores (free since Aug 2025)
Not our product — use Steam's.
"How many players / sales by country (estimates)?"
Gamalytic (from $25/mo), GameDiscoverCo Plus ($19/mo)
Not our product — we provide no quantitative estimates.
"What are Japanese players saying, by topic, with quotes?"
Read them yourself through auto-translate + DIY AI summary
This is the product: classified counts + quoted examples + Japan-specific vs global separation
"What happened in Japan around my game this month (media, creators, calendar)?"
Manual Japanese-language searching
This is the product: Japanese sources, read in Japanese, summarised in English
"So what should I do next, specifically?"
Your own judgement from the above
At most 3 moves, each tied to a finding, each with an execution-cost note
If the free rows answer your question, do not pay us. Radar only makes sense if the last three rows are the ones you cannot cover.
Pricing (shown before any sign-up — deliberately)
Solo
Studio
Price
$49 / month
$99 / month
Titles
1 title
Up to 5 titles
Report
1 monthly report (contents above)
1 monthly report per title
Annual option (if launched)
2 months free on annual billing
2 months free on annual billing
Prices are in USD, before any taxes that may apply at launch.
No custom work at either tier. No extra titles mid-month, no rush delivery, no bespoke topics. The format is fixed so the price can stay at this level.
Launch is not guaranteed. If Radar does not launch, waitlist sign-ups receive one email saying so, and no charge ever occurs.
Paid pilot — try one report before any subscription exists
$49, one-time. One monthly report for one title. No auto-renew, no subscription created.
After delivery, we ask once — and only once — whether you would continue at $49/month. Your answer (yes / no / why) is the product research. A "no" with a reason is a useful result for us and costs you nothing extra.
Pilot slots are limited by production capacity (each report is produced and checked per title). If slots are full, we will say so and offer the waitlist instead of taking your money.
How it works
Join the waitlist (or apply for a pilot) with your game's Steam URL. The Steam URL is required — it is how we keep the list to teams with a real title to report on.
Waitlist members receive: the full sample, and one launch/pilot invitation email. No newsletter, no drip campaign.
Pilot applicants receive a confirmation, a delivery date, and payment instructions for the one-off $49. Delivery is a document by email.
FAQ
Q. Why monthly? My game only needs Japan attention at launch. A. That may be true — and if your need is one-off, take the pilot, not a subscription. Monthly exists because the full-sample data shows Japanese topics move (PEAK's frustration centre shifted from bugs/language to Deep Forest balance). If your title is dormant, we would rather you churn honestly than pay for empty reports.
Q. Can't I just use ChatGPT / Claude on my Japanese reviews myself? A. You can, and for a rough read you should. What DIY does not give you is the standing classification scheme (comparable month to month), the Japan-specific vs global separation, Japanese media/creator context, and someone whose job is to notice the Japanese calendar event before it passes. If DIY covers your need, do not pay us.
Q. Do you cover X/Twitter? A. Not comprehensively — see Scope. Anyone promising full Japanese X monitoring at this price is either violating platform terms or losing money on API fees. We spot-check manually around specific events and label it as such.
Q. Which games is this for? A. Small Steam teams and small publishers with a shipped title that has meaningful Japanese interest (Japanese reviews, Japanese creator play, or a Japanese localisation) and no in-house Japanese staff. If your game has almost no Japanese footprint, a report will be thin and we will tell you that before a pilot, not after.
Q. What languages / genres? A. Reports are in English. Source material is Japanese. Genre is not restricted, but our samples are co-op titles because that is where Japanese creator-driven demand is easiest to demonstrate. Other genres are welcome on the waitlist; pilot acceptance depends on whether we can produce a non-thin report for your title.
Q. Refunds? A. Pilot: if we cannot deliver a non-thin report for your title, we refund the $49 in full rather than pad it. Subscription terms will be published before any subscription is sold.
Q. Who makes this? A. A small Japan-based research operation using AI-assisted collection and classification with human-checked interpretation. Collection uses official APIs and public sources within their terms (Steam reviews via Steam's official Web API). We state our sources and snapshot dates in every report, and we label what we could not verify.
Q. Will you do a custom report / call / extra analysis if I pay more? A. No. This is deliberate: fixed format is what makes $49 possible and keeps delivery reliable.
Join the waitlist
You have seen the prices ($49 Solo / $99 Studio) and the scope above. Signing up below therefore means: "at those prices, for that scope, I want the full sample and a launch/pilot invitation." That is the only thing a waitlist sign-up claims — we do not treat it as a purchase.