Why Domain Authority Doesn't Predict SERP Position in Japan

Why This Study Exists

The conventional SEO narrative — "higher Domain Authority predicts higher SERP position" — is widely cited but increasingly contested by practitioner observation. A search result page for "neighborhood ramen recommendations" looks structurally different from one for "advanced cardiology referral," and not just because of content topic. The visible structure of which sites occupy positions 1-10 differs in ways that don't reduce to a single Domain Authority gradient.

This study tests that practitioner observation against data. We analyzed 11,116 Japanese SERP results across 24 industries, indexing 1,720 unique domains and their DR values, to ask: when does Domain Authority actually predict ranking, and when does it not — and what explains the difference?

JFSC's interest in the question is not abstract. As an independent M&A advisor working with Japanese SME owners, the structure of search visibility directly shapes which industries offer standalone scaling pathways versus which industries push owners toward consolidation or platform-mediated exit. The data informs which exit pattern is structurally available for each industry — not what the owner should do, but what the market structure permits.

The Core Finding — DR Does Not Predict SERP Uniformly

Overall Pearson correlation between Domain Authority (Ahrefs DR) and SERP rank across all 11,116 results: ρ = −0.015. That is statistically indistinguishable from zero.

However, stratified by industry × user-intent layer (50 strata), the correlation varies from ρ = −0.315 to ρ = +0.283 — a 0.6-point spread that is anything but uniform:

Two industries (manufacturing, IT) show DR positively correlating with SERP rank — the conventional SEO model holds. Three industries (food service, cosmetic medical, elderly care) show DR inverse-correlating with rank — heavy SEO investment by mid-tier domains does not produce ranking gains, and may actively suppress them. The remaining industries cluster around zero correlation.

Three SERP Oligopoly Typologies

Across 24 industries, three structural patterns emerge:

Typology 1 — Fragmented Competition (gunyuu kakkyo)

No single dominant player; many mid-tier domains compete on roughly equal footing. DR positively correlates with rank because there is no structural alternative to building organic authority. Examples: manufacturing, mid-B2B services, specialty trading.

Typology 2 — Mid-Tier Survival (chuuken seizon)

Platform layer present but mid-tier domains co-exist; DR weakly correlates with rank. Mid-tier operators can rank but face competition from platform-adjacent sites. Examples: cosmetic medical, fitness, beauty salon.

Typology 3 — PLATFORM Proxy (PLATFORM daikou)

One or two dominant platforms absorb the majority of top-rank positions; individual operators surface only via platform intermediation. DR-rank correlation negative or zero. Examples: food service (Tabelog, Gurunavi, Hot Pepper); elderly care (kaigo platforms); accommodation (Booking.com, Rakuten Travel, Jalan).

The typology is not a quality judgment — it is a structural observation about how monetization in the industry shapes who gets surface visibility. Industries with strong platform intermediation systematically push individual operators toward platform partnership; industries without it create more direct content competition.

Industry-Level Fingerprints

PLATFORM Duality and Intermediation Layer

PLATFORM duality refers to the dual function of dominant intermediation platforms:

1. As the primary visibility layer for end-user search queries
2. As the dominant economic counterparty for individual operators

In industries with strong PLATFORM duality, individual operators face structural margin compression because their visibility depends on platform inclusion, and the platform extracts a meaningful share of the resulting transaction value. Food service operators pay 8-15% to Tabelog / Gurunavi-equivalent commission structures. Elderly care operators face similar dependency, mediated through public-sector facility-listing systems.

The intermediation layer can be decomposed into 4 tiers:

1. Direct sales channel (the operator's own website, walk-in, referral)
2. Comparison and review aggregators (kakaku.com, Tabelog reviews)
3. Booking / transaction platforms (Booking.com, Rakuten Travel)
4. Discovery and search (Google search itself + map_pack)

The structural question is which tier captures economic value. In PLATFORM-duality industries, tiers 2-4 capture more value than tier 1; in fragmented industries, tier 1 retains more value but faces direct competition rather than platform mediation.

The map_pack Effect — Elderly Care Hostage-Structure Reversal

Initial hypothesis: in elderly care, platform dominance creates a "hostage structure" where operators are forced into platform-listed visibility, producing concentrated review volume on platform listings.

Data observation: review volume is concentrated on platforms, but the causation direction is reversed. The platforms attract reviews because they are the SERP top results — not vice versa. The map_pack (Google local pack) shows HHI concentration of +187 to +235 points in elderly care relative to base industries, indicating that local-search algorithmic preferences reinforce platform dominance independent of operator-side review-accumulation efforts.

AIO Adoption — Japanese Mid-Market Lag

Per the companion study (ai-aio-30company-research-2026), Japanese listed mid-market companies show a wait-and-see posture on AI Overview (AIO) optimization. Of 30 surveyed companies across Japan-US comparison:

• US listed peers showed structural markup adoption (Schema.org Organization, Article, FAQ; entity disambiguation; Wikipedia presence) at 73%
• Japanese listed peers showed equivalent adoption at 18%
• Cross-LLM citation behavior: Gemini Deep Think suppresses low-quality citation by 243% vs Flash baseline; Claude and ChatGPT show modest +10% to +14% increase

The structural gap creates an asymmetric early-mover opportunity for Japanese companies that invest in AIO-optimized content presence — capturing AIO-citation share at lower cost than will be available once peer adoption catches up. This is one of the few SEO investment areas where the conventional "build before competitors" logic still applies in Japanese B2B search.

M&A Intermediation Implications

Two distinct M&A strategy patterns emerge from the SERP structure data:

Exit-Driven Path

In PLATFORM-dominated industries (food service, elderly care, accommodation), individual operators have reduced standalone scaling potential because the SERP top positions are structurally inaccessible. The structurally rational exit is sale to a consolidator with platform leverage, or sale to a platform-adjacent acquirer (regional aggregator, vertical specialist). The owner's negotiating leverage is operator-quality and customer base — not domain authority or SEO position.

Roll-Up-Driven Path

In fragmented competition industries (manufacturing, B2B services, specialty trading), consolidators can build margin by aggregating mid-tier operators to internalize buying power, customer reach, and SEO authority. The structurally rational acquirer is a strategic consolidator pursuing operational scale, often with the explicit goal of building post-merger SERP dominance through consolidated content authority.

Methodology, Limitations, and Self-Retractions

Methodology Summary

• 11,116 SERP results captured from Japanese Google (google.co.jp), May 2026 snapshot
• 1,720 unique domains indexed; Ahrefs DR captured per domain
• 24 industries × 5 user-intent layers (informational / commercial / navigational / transactional / local) = 50 strata for correlation analysis
• Pearson and Spearman correlations both computed; reported figure is Pearson
• map_pack HHI computed independently for industries with map_pack-prominent intent layers

Limitations

1. Japan-only corpus; not validated against English or other-language SERP behavior
2. Sample reflects May 2026 snapshot — algorithm updates may shift the picture; intended as periodic re-survey
3. DR measurement is Ahrefs-based; alternative DR metrics (Moz DA, Semrush AS) may produce different correlation patterns
4. 24-industry coverage is broad but not exhaustive of Japanese economy
5. SERP capture from a single geographic origin; mobile / desktop blending; personalization controlled but not fully isolated

Self-Retractions and Method Disclosures

One industry tabulation (AIO aggregation for Japanese mid-market) had an agent-execution error in the initial run, producing artifactual results. The error was identified during cross-check; the AIO tabulation was retracted and re-analyzed. The published figures reflect the corrected run. The retraction is disclosed here in the spirit of method transparency.

Falsification Conditions

The core finding (DR does not uniformly predict SERP) would be falsified by demonstrating that, under matched query-intent and matched evaluator conditions, DR alone explains over 30% of SERP-rank variance across all 24 industries. Independent replication is invited.

Frequently Asked Questions

About the Author

Yuichi Igarashi — Founder & CEO, Japan Financial Strategy Center (JFSC). Graduate of Kyoto University Faculty of Economics. Prior experience at Sompo Japan Insurance Inc. (corporate risk and legal practice) and a Tokyo Stock Exchange–listed M&A intermediary firm. Founded JFSC in 2020. JFSC's research portfolio includes the "Induced Intermediation" series examining how information-asymmetry intermediaries interact with search-engine quality signals; the SERP Industry Structure study reported here is one component of that broader portfolio. Registered M&A Support Organization (SME Agency).

Read the founder's full message · About JFSC · Why JFSC