Tabelog Ratings Explained: Why 3.5 Is Elite, and What the 2026 Court Ruling Settled

Last updated July 22, 2026 — includes the Supreme Court of Japan’s final ruling of March 5, 2026, and Tabelog’s current published scoring rules.

The Tabelog rating system is the most misread number in Japanese dining: scores are not averages, they’re weighted by each reviewer’s influence, and by Tabelog’s own published distribution only 3% of restaurants score 3.5 or higher — which is why a “low” 3.2 is often a very good restaurant, and why comparing Tabelog to Google’s ratings misleads both diners and owners.

Key takeaways

• 3.5 on Tabelog is elite: Tabelog’s published distribution puts 3.50–3.99 in the top 3% of restaurants and 4.0+ in the top 0.07%. Read 3.0–3.4 as “normal good,” not “mediocre.”
• Scores are influence-weighted, not averaged — a reviewer’s weight depends on their track record in that cuisine genre; suspected fake or conflicted accounts count for zero.
• Money doesn’t move the score. Tabelog states paid plans have no effect on ratings — and after a seven-year lawsuit over its 2019 algorithm change, Japan’s Supreme Court dismissed the final appeal on March 5, 2026, leaving Tabelog’s win intact.
• For restaurant owners, the actionable channel is Google: 86% of Japanese diners use Google to find restaurants vs 61% for gourmet platforms (2022 industry survey) — and unlike Tabelog’s algorithm, Google visibility responds to systematic review collection, which is what ReputeMap automates.

How does Tabelog’s rating system actually work?

Tabelog says it plainly on its own help pages: the score is not a simple average of user ratings. Instead, every reviewer carries an influence weight (影響度), and the score is computed from ratings scaled by that weight. Three official rules shape everything:

1. Influence is per-genre. A reviewer with a long, consistent history of ramen reviews carries real weight for ramen shops — and near-none for a patisserie they review once. Expertise is measured where it was earned.
2. Zero-influence accounts exist by design. Users flagged for possible stealth marketing, restaurant staff, and anyone with a stake in the outcome are assigned no influence at all — their stars display, but don’t move the number.
3. Volume matters. A handful of glowing reviews can’t produce a high score; the algorithm requires breadth before it trusts a number.

Scores recalculate twice a month — the first and third Tuesday — which is why a restaurant’s number moves in steps, not daily drips.

The result is a deliberately compressed curve. Most listed restaurants live between 3.0 and 3.4; crossing 3.5 puts you in the top 3%, and 4.0 is rarer than a Michelin star. Tourists who filter Japan by “4+ stars” out of Google habit filter out nearly the entire country.

Does paying Tabelog raise a score? Seven years of courtrooms say no

This is the question that spawned a thousand forum threads — and, unusually, it has a definitive legal answer.

Tabelog’s official position has always been that paid restaurant services have zero effect on scores. The stress test came from the “3.8 problem” era: after a May 2019 algorithm change, the Korean-barbecue chain KollaBo saw scores across ~20 outlets drop by about 0.2 points on average (0.45 at worst), and its operator Hanryumura sued, alleging the algorithm secretly demoted chain restaurants to pressure them into higher-paying plans.

The case ran the full ladder of Japanese courts:

Date	Court	Outcome
June 2022	Tokyo District Court	Ruled the uniform chain-store demotion an “abuse of superior bargaining position” under the Antimonopoly Act; awarded Hanryumura ¥38.4M
January 2024	Tokyo High Court	Reversed in full — found the algorithm update justified as a measure against review distortion
March 5, 2026	Supreme Court of Japan	Dismissed the final appeal, leaving the High Court ruling final and binding

Two honest readings coexist: the courts ultimately sided with Tabelog’s right to change its algorithm against manipulation — and the seven-year fight is exactly why many Japanese owners remain wary of a score they can neither audit nor influence. Both things are true, and both matter for strategy.

Tabelog vs Google Maps: which number matters in 2026?

For diners, the practical answer travelers have converged on: Tabelog tells you if the food is good (on its compressed scale), Google tells you where things are and what’s open. The same restaurant routinely shows a 3.3 on one and 4.6 on the other — that’s two different measuring systems, not a contradiction.

For restaurant owners, the discovery data is more one-sided than most assume: in a 1,100-respondent TableCheck survey, 86.1% of Japanese diners used Google to find restaurants versus 61.3% for the gourmet platforms — and the same research arc showed distrust of gourmet-site ratings rising (26.1% → 30.4%, 2020→2022). Japan is a market where Google quietly became the primary discovery layer while everyone was watching the incumbent (the full Japan picture is in our Asia country guide).

If you run a restaurant in Japan: what this means practically

The strategic asymmetry is the whole story:

• Tabelog’s score resists direct effort by design. Influence weighting means you can’t ask your way to 3.5; paid plans don’t touch the number (officially and now judicially); the right move is hygiene — accurate listing, good photos, honest service over time.
• Google visibility responds to systematic work. Complete Japanese-language profile, fresh photos, and above all a steady cadence of reviews with owner replies in the reviewer’s language — the same playbook as anywhere, applied to the channel 86% of your customers actually search on. Start with our profile setup checklist.

That weekly cadence — asking every table honestly, replying to every review in Japanese or English — is exactly the part that dies in a busy service week, and it’s what ReputeMap automates for $29–49/month: requests by email, QR or link, every review in one inbox with an AI-drafted reply you approve, and negative-review alerts within minutes. Tabelog will keep its own counsel; your Google presence is yours to build.

Sources: Tabelog’s official score documentation (tabelog.com/help/score); Japan Today and court-coverage reporting on Hanryumura v. Kakaku.com (2022–2026); Greenberg Traurig Competition Currents, June 2026 (Supreme Court dismissal, March 5, 2026); TableCheck diner survey via ITmedia (2022). Consumer statistics: our verified statistics hub.