More Reviews, More Clients: How Google Review Volume, Not Star Ratings, Drives Organic Traffic, Brand Demand, and AI Visibility Across 1,000 US Personal Injury Law Firms

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

Google reviews have quietly become the most regulated, most contested, and most multiply-consumed asset in local search. Between July 2025 and April 2026, Google merged Local Services Ads reviews into the Google Business Profile review stream, replaced the Google Screened badge with a single Verified checkmark, and rewrote its Maps rating-manipulation policy to prohibit review quotas, contests, and per-staff incentives. The same review corpus now feeds four surfaces at once: the local pack, the Business Panel, Local Services Ads, and the AI answer layer. This paper, the ninth in a series on entity-completeness signals in the top-ranking US personal injury (PI) vertical, is the first large-N audit of that review economy. For the series' 1,000 top-ranking US PI law firm sample we join the Google Business Panel review signals (review count, star rating, office-place count) captured in the August 2026 Business Panel audit to the organic-traffic, brand-search-demand, domain-authority, Knowledge Panel, and per-assistant AI-citation signals established in the prior papers. Five findings define the review economy. First, review capital is deeply unequal: the 732 rated firms hold 192,590 reviews, the median firm holds 88, the top decile of firms holds 56.8% of the entire stock, and the Gini coefficient is 0.70, the same winner-take-most shape the series documented for AI citations. Second, the star scale has stopped discriminating: 80.6% of rated firms score 4.5 or higher and 47.5% score 4.9 or higher, and the rating predicts no outcome (traffic rho = -0.06, brand demand rho = 0.03, AI mentions rho = -0.07, all non-significant). Third, review volume predicts every outcome the rating fails to predict: rho = 0.40 with organic traffic, 0.39 with brand search volume, 0.39 with total AI mentions, and 0.41 with ChatGPT mentions, and firms with ten or more reviews carry 6.4x the median organic traffic of firms below ten (p < 0.001). Fourth, the perfection band underperforms: firms rated 4.9-5.0 hold more median reviews than firms rated 4.5-4.8 (135 vs 95) yet draw 40% less median traffic (260 vs 430, p < 0.001), a pattern we name the perfection plateau. Fifth, the office footprint is a null: multi-location firms show no traffic or AI-visibility advantage over single-location firms (p = 0.50 and p = 0.89), while reviews-per-location tracks outcomes closely, so the corpus, not the pin count, carries the signal. In regression, review volume remains a significant predictor of traffic net of authority, brand demand, and both Knowledge Panel classes (p = 0.024), and review volume is the strongest standardized correlate of brand search demand in its model (beta = 0.29, ahead of authority at 0.17), placing reviews upstream of the Brand Demand Flywheel documented earlier in the series. A deliberately constructed composite, the Review Capital Index (RCI), underperforms raw review volume against every outcome, which is itself the finding: review capital is one-dimensional, and the dimension is volume. The paper closes with Review Capital Optimization (RCO), a compliance-native playbook for the post-April-2026 policy regime.