{"id":114,"date":"2026-04-02T10:54:02","date_gmt":"2026-04-02T10:54:02","guid":{"rendered":"https:\/\/promptlocal.ai\/blog\/?p=114"},"modified":"2026-04-03T14:00:42","modified_gmt":"2026-04-03T14:00:42","slug":"personal-injury-lawyers-ai-results","status":"publish","type":"post","link":"https:\/\/promptlocal.ai\/blog\/personal-injury-lawyers-ai-results\/","title":{"rendered":"What Does AI Think About Personal Injury Lawyers in Las Vegas?"},"content":{"rendered":"\n<link href=\"https:\/\/fonts.googleapis.com\/css2?family=DM+Serif+Display&#038;family=Space+Mono:wght@400;700&#038;family=DM+Sans:wght@400;500;600;700&#038;display=swap\" rel=\"stylesheet\">\n<style>\n\/* ===== MOBILE RESPONSIVE IMPROVEMENTS ===== *\/\n@media (max-width: 768px){\n\n  .container{\n    max-width: 100%;\n    padding: 0 16px;\n  }\n\n  .hero{\n    padding: 50px 0 40px;\n  }\n\n  .hero h1{\n    font-size: 28px;\n    line-height: 1.25;\n  }\n\np,span,div,td,th{\nfont-size:14px !important;\n}\n.hero-sub{\nfont-size:14px !important;\n}\n.stat-num{\nfont-size:22px !important;\n}\n\n  section{\n    padding: 40px 0;\n  }\n\n  section h2{\n    font-size: 22px;\n  }\n\n  .section-sub{\n    font-size: 14px;\n  }\n\n  .stats-grid{\n    grid-template-columns: repeat(2, 1fr);\n    gap: 10px;\n  }\n\n  .stat-card{\n    padding: 16px;\n  }\n\n  .stat-num{\n    font-size: 24px;\n  }\n\n  .stat-label{\n    font-size: 14px;\n  }\n\n  \/* BAR CHART FIX *\/\n  .bar-row{\n    flex-direction: column;\n    align-items: flex-start;\n    gap: 4px;\n    margin-bottom: 12px;\n  }\n\n  .bar-label{\n    width: 100%;\n    text-align: left;\n    font-size: 11px;\n    padding-right: 0;\n  }\n\n  .bar-track{\n    width: 100%;\n    height: 22px;\n  }\n\n  .bar-meta{\n    width: 100%;\n    padding-left: 0;\n    font-size: 10px;\n  }\n\n  \/* TABLE SCROLL *\/\n  table.dt{\n    display: block;\n    overflow-x: auto;\n    white-space: nowrap;\n  }\n\n  .dt th, .dt td{\n    font-size: 11px;\n    padding: 8px 6px;\n  }\n\n  \/* OVERLAP GRID *\/\n  .overlap-grid{\n    grid-template-columns: 1fr;\n  }\n\n  \/* DONUT SECTION *\/\n  .donut-wrap{\n    flex-direction: column;\n    align-items: center;\n    gap: 16px;\n  }\n\n  .legend{\n    width: 100%;\n  }\n\n  .legend-item{\n    justify-content: flex-start;\n  }\n\n  \/* NOMENCLATURE *\/\n  .nom-grid{\n    grid-template-columns: 1fr;\n    gap: 6px 0;\n  }\n\n  \/* CALLOUTS *\/\n  .callout{\n    padding: 16px 18px;\n    font-size: 14px;\n  }\n\n  .example-box{\n    padding: 14px 16px;\n    font-size: 13px;\n  }\n\n  \/* SVG CHARTS RESPONSIVE *\/\n  svg{\n    width: 100% !important;\n    height: auto !important;\n  }\n\n  \/* FOOTER *\/\n  .footer{\n    font-size: 12px;\n    padding: 30px 0;\n  }\n}\n\n\/* EXTRA SMALL DEVICES *\/\n@media (max-width: 480px){\n\n  .hero h1{\n    font-size: 24px;\n  }\n\n  .hero-sub{\n    font-size: 13px;\n  }\np,span,div,td,th{\nfont-size:13px !important;\n}\n  .stat-num{\n    font-size: 24px;\n  }\n\n  section h2{\n    font-size: 20px;\n  }\n\n  .tag{\n    font-size: 9px;\n    padding: 2px 10px;\n  }\n\n  .bar-track{\n    height: 18px;\n  }\n\n  .bar-fill{\n    font-size: 9px;\n  }\n}\n:root{--bg:#0a0a0f;--surface:#0f172a;--surface2:#1a1a28;--border:#2a2a3a;--text:#e4e4ef;--dim:#8888a0;--accent:#ff6b35;--accent2:#00d4aa;--accent3:#6366f1;--accent4:#f59e0b;--red:#ef4444;--green:#22c55e}\n*{margin:0;padding:0;box-sizing:border-box}\nbody{font-family:var(--theme-font-family);background:var(--bg);color:var(--text);line-height:1.7}\n.container{max-width:90%;margin:0 auto;padding:0 24px}\n.hero{padding:0 0 60px;text-align:center;border-bottom:1px solid var(--border)}\n.hero-label{font-family:var(--theme-font-family);font-size:12px;letter-spacing:3px;text-transform:uppercase;color:var(--accent);margin-bottom:20px}\n.hero h1{font-family:var(--theme-font-family);font-size:clamp(32px,5vw,52px);line-height:1.15;color:#fff;max-width:900px;margin:0 auto 24px}\np,span,div,td,th{font-size:16px !important}\n.hero-sub{font-size:16px;color:var(--dim);max-width:750px;margin:0 auto}\n.stat-num{font-size:30px !important}\n.hero-sub strong{color:var(--accent)}\n.stats-grid{display:grid;grid-template-columns:repeat(auto-fit,minmax(155px,1fr));gap:14px;margin:48px 0}\n.stat-card{background:var(--surface);border:1px solid var(--border);border-radius:12px;padding:22px;text-align:center}\n.stat-num{font-family:var(--theme-font-family);font-size:34px;font-weight:700;color:var(--accent)}\n.stat-label{font-size:12px;color:var(--dim);margin-top:4px}\nsection{padding:56px 0;border-bottom:1px solid var(--border)}\nsection h2{font-family:var(--theme-font-family);font-size:30px;color:#fff;margin-bottom:8px}\n.section-sub{color:#ffffff;margin-bottom:28px;font-size:15px}\n.tag{display:inline-block;font-family:var(--theme-font-family);font-size:11px;letter-spacing:2px;text-transform:uppercase;color:var(--accent);background:rgba(255,107,53,.1);border:1px solid rgba(255,107,53,.3);padding:3px 12px;border-radius:100px;margin-bottom:14px}\n.callout{    background: var(--surface);\n    border-left: 4px solid var(--accent);\n    border-radius: 0 10px 10px 0;\n    padding: 18px 22px;\n    margin: 16px 0;\n    font-size: 14px;}\n.callout strong{color:var(--accent)}\n.callout-green{border-left-color:var(--accent2)}\n.callout-green strong{color:var(--accent2)}\n.callout-purple{border-left-color:var(--accent3)}\n.callout-purple strong{color:var(--accent3)}\n.example-box{background:var(--surface2);border:1px solid var(--border);border-radius:10px;padding:18px 22px;margin:16px 0;font-size:14px}\n.example-box .ex-label{font-family:var(--theme-font-family);font-size:10px;letter-spacing:1.5px;text-transform:uppercase;color:var(--accent4);margin-bottom:6px}\n.bar-chart{margin:20px 0;text-align:center;}\n.bar-row{display:flex;align-items:center;margin-bottom:5px}\n.bar-label{width:240px;font-size:12px;text-align:right;padding-right:12px;white-space:nowrap;overflow:hidden;text-overflow:ellipsis;flex-shrink:0}\n.bar-track{flex:1;height:26px;background:var(--surface2);border-radius:4px;overflow:hidden}\n.bar-fill{height:100%;border-radius:4px;display:flex;align-items:center;padding-left:8px;font-family:var(--theme-font-family);font-size:11px;color:#fff;min-width:36px}\n.bar-meta{width:140px;font-size:12px;color:var(--dim);padding-left:10px;font-family:var(--theme-font-family);flex-shrink:0;text-align: start;}\ntable.dt{width:100%;border-collapse:collapse;margin:20px 0}\n.dt th{font-family:var(--theme-font-family);font-size:10px;letter-spacing:1px;text-transform:uppercase;color:var(--dim);border-bottom:2px solid var(--border);padding:10px 8px;text-align:left}\n.dt td{padding:9px 8px;border-bottom:1px solid var(--border);font-size:13px}\n.dt tr:hover td{background:var(--surface)}\n.dt .n{font-family:var(--theme-font-family);text-align:right}\n.overlap-grid{display:grid;grid-template-columns:repeat(2, 1fr);gap:16px;margin:20px 0}\n.ov-card{background:var(--surface);border:1px solid var(--border);border-radius:12px;padding:18px}\n.ov-card h4{font-size:14px;margin-bottom:3px;color:#fff}\n.ov-card .sub{font-size:11px;color:var(--dim);margin-bottom:10px}\n.ov-bar{height:7px;background:var(--surface2);border-radius:4px;margin-bottom:6px;overflow:hidden}\n.ov-fill{height:100%;border-radius:4px}\n.ov-pct{font-family:var(--theme-font-family);font-size:26px;font-weight:700}\n.donut-wrap{display:flex;align-items:center;gap:28px;margin:20px 0;flex-wrap:wrap;justify-content:center}\n.legend{display:flex;flex-direction:column;gap:10px}\n.legend-item{display:flex;align-items:center;gap:10px}\n.legend-dot{width:13px;height:13px;border-radius:3px;flex-shrink:0}\n.legend-pct{font-family:var(--theme-font-family);font-weight:700;font-size:14px}\n.legend-label{font-size:13px;color:var(--dim)}\n.prose{font-size:15px;line-height:1.8;}\n.prose p{margin-bottom:14px}\n.hl{color:var(--accent);font-weight:600}\n.hl-g{color:var(--accent2);font-weight:600}\n.hl-p{color:var(--accent3);font-weight:600}\n.hl-y{color:var(--accent4);font-weight:600}\n.nomenclature{background:var(--surface);border:1px solid var(--border);border-radius:12px;padding:24px 28px;margin:28px 0}\n.nomenclature h3{font-size:16px;color:#fff;margin-bottom:12px}\n.nom-grid{display:grid;grid-template-columns:1fr 1fr;gap:8px 24px;font-size:13px}\n.nom-grid .lbl{color:var(--dim)}\n.limitation{background:rgba(239,68,68,.06);border:1px solid rgba(239,68,68,.2);border-radius:10px;padding:18px 22px;margin:16px 0;font-size:14px}\n.limitation .lim-title{color:var(--red);font-weight:600;margin-bottom:4px}\n.footer{padding:40px 0;text-align:center;color:var(--dim);font-size:13px}\n.footer a{color:var(--accent);text-decoration:none}\n@media(max-width:768px){.bar-label{width:110px;font-size:10px}.bar-meta{width:90px;font-size:10px}.stats-grid{grid-template-columns:repeat(1,1fr)}.nom-grid{grid-template-columns:1fr}}\n\ntable.dt{\n  width: 100%;\n  border-collapse: separate;\n  border-spacing: 0;\n  border: 1px solid var(--border);\n  border-radius: 12px;\n  overflow: hidden;\n  background: var(--surface);\n}\n\n.dt th{\n  background: var(--surface2);\n  border-bottom: 1px solid var(--border);\n}\n\n.dt td{\n  border-bottom: 1px solid var(--border);\n}\n\n.dt tr:last-child td{\n  border-bottom: none;\n}\n\ntable, th, td {\n    border-width: var(--theme-table-border-width, 0px);\n}\n\n\/* ============================= *\/\n\/* MOBILE ONLY (<=768px) *\/\n\/* ============================= *\/\n@media (max-width:768px){\n\n  \/* CONTAINER *\/\n  .container{\n    max-width:100%;\n    padding:0 14px;\n  }\n\n  \/* TYPOGRAPHY *\/\n  body{\n    font-size:14px;\n    line-height:1.6;\n  }\n\n  p{\n    font-size:14px;\n  }\n\n  .section-sub{\n    font-size:13px;\n  }\n\n  \/* HERO *\/\n  .hero{\n    padding:40px 0 30px;\n  }\n\n  .hero h1{\n    font-size:26px;\n    line-height:1.3;\n  }\n\n  .hero-sub{\n    font-size:14px;\n  }\n\n  \/* SECTIONS *\/\n  section{\n    padding:36px 0;\n  }\n\n  section h2{\n    font-size:22px;\n  }\n\n  \/* STATS *\/\n  .stats-grid{\n    grid-template-columns:repeat(2,1fr);\n    gap:10px;\n  }\n\n  .stat-card{\n    padding:16px;\n  }\n\n  .stat-num{\n    font-size:24px;\n  }\n\n  .stat-label{\n    font-size:12px;\n  }\n\n  \/* BAR CHART *\/\n  .bar-row{\n    display:flex;\n    flex-direction:column;\n    align-items:stretch;\n    gap:6px;\n    margin-bottom:12px;\n  }\n\n  .bar-label{\n    width:100%;\n    text-align:left;\n    font-size:12px;\n    padding-right:0;\n  }\n\n  .bar-track{\n    width:100%;\n    height:20px;\n  }\n\n  .bar-meta{\n    width:100%;\n    font-size:11px;\n    padding-left:0;\n  }\n\n  \/* TABLE SCROLL *\/\n  table.dt{\n    display:block;\n    overflow-x:auto;\n    white-space:nowrap;\n  }\n\n  .dt th, .dt td{\n    font-size:12px;\n    padding:6px;\n  }\n\n  \/* GRID STACK *\/\n  .overlap-grid{\n    grid-template-columns:1fr;\n  }\n\n  \/* DONUT *\/\n  .donut-wrap{\n    flex-direction:column;\n    align-items:center;\n    text-align:center;\n    gap:16px;\n  }\n\n  .legend{\n    width:100%;\n    text-align: left;\n  }\n\n  .legend-item{\n    justify-content:flex-start;\n  }\n\n  \/* CALLOUTS *\/\n  .callout{\n    font-size:14px;\n    line-height:1.6;\n    padding:16px;\n  }\n\n  .example-box{\n    font-size:13px;\n    padding:14px;\n  }\n\n  \/* SVG FIX *\/\n  svg{\n    width:100% !important;\n    height:auto !important;\n    max-width:100%;\n  }\n\n  \/* SCROLLABLE CHARTS *\/\n  .chart-scroll{\n    width:100%;\n    overflow-x:auto;\n  }\n\n  \/* FOOTER *\/\n  .footer{\n    font-size:12px;\n    padding:30px 0;\n  }\n}\n\n\n\/* ============================= *\/\n\/* EXTRA SMALL (<=480px) *\/\n\/* ============================= *\/\n@media (max-width:480px){\n\n  .hero h1{\n    font-size:22px;\n  }\n\n  .hero-sub{\n    font-size:13px;\n  }\n\n  .stats-grid{\n    grid-template-columns:1fr;\n  }\n\n  .tag{\n    font-size:9px;\n    letter-spacing:1px;\n    padding:10px 24px;\n  }\n\n  .bar-track{\n    height:18px;\n  }\n\n  .bar-fill{\n    font-size:9px;\n  }\n}\n\n@media (max-width:768px){\n\n  .tag{\n    display:inline-block;\n    font-size:10px;\n    letter-spacing:1.2px;\n    padding:10px 20px;\n    border-radius:999px;\n    line-height:1.2;\n    white-space:normal;       \/* allows wrapping *\/\n    max-width:100%;\n    word-break:break-word;\n  }\n\n}\n<\/style>\n\n\n<div class=\"container hero\">\n<div class=\"container\">\n<div class=\"hero-label\">PromptLocal.ai Research Report<\/div>\n<h1>What Does AI Think About Personal Injury Lawyers in Las Vegas?<\/h1>\n<p class=\"hero-sub\">We asked <strong>3 AI models<\/strong> the same <strong>124 questions<\/strong> about personal injury lawyers across <strong>3 channels<\/strong> \u2014 prompting AI interfaces directly, querying APIs, and Google AI Mode. Here's what they got right, and where they wildly disagree.<\/p>\n<\/div>\n<\/div>\n\n<div class=\"container\">\n\n<!-- NOMENCLATURE -->\n<div class=\"nomenclature\">\n<h3>How We Tested: 3-Channel Framework<\/h3>\n<div class=\"nom-grid\">\n<div class=\"lbl\">Channel 1 \u2014 AI interfaces<\/div><div>Prompting AI interfaces directly (ChatGPT, Gemini, Perplexity & Google AI Mode Web UI)<\/div>\n<div class=\"lbl\">Channel 2 \u2014 API<\/div><div>Same models queried programmatically via their respective APIs<\/div>\n<div class=\"lbl\">Channel 3 \u2014 Google Search<\/div><div>Local Pack + Organic results<\/div>\n<div class=\"lbl\">Web Search ON<\/div><div>AI has live internet access (browsing, grounding, citations)<\/div>\n<div class=\"lbl\">Web Search OFF<\/div><div>AI relies only on training data \u2014 no live web access<\/div>\n<\/div>\n<\/div>\n\n<div class=\"stats-grid\">\n<div class=\"stat-card\"><div class=\"stat-num\">~1,860<\/div><div class=\"stat-label\">Total AI Queries<\/div><\/div>\n<div class=\"stat-card\"><div class=\"stat-num\">1,267<\/div><div class=\"stat-label\">Unique Firms Mentioned<\/div><\/div>\n<div class=\"stat-card\"><div class=\"stat-num\">10,749<\/div><div class=\"stat-label\">Total Firms Mentioned<\/div><\/div>\n<div class=\"stat-card\"><div class=\"stat-num\">15<\/div><div class=\"stat-label\">Channels Tested<\/div><\/div>\n<div class=\"stat-card\"><div class=\"stat-num\">66.9%<\/div><div class=\"stat-label\">Firms Known by 1 Channel Only<\/div><\/div>\n<div class=\"stat-card\"><div class=\"stat-num\">6,416<\/div><div class=\"stat-label\">Citation URLs Extracted<\/div><\/div>\n<\/div>\n\n<!-- ===== FINDING 1: LEADERBOARD ===== -->\n<section>\n<div class=\"tag\">Finding 01 \u00b7 Mention Frequency + Rank + Share of Voice<\/div>\n<h2>The AI Visibility Leaderboard<\/h2>\n<p class=\"section-sub\">Out of 1,267 unique law firms, the top 5 control 26% of all recommendations. There's a clear AI canon \u2014 and most firms aren't in it.<\/p>\n<div class=\"bar-chart\" id=\"top15chart\"><\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 Rank #1 \u2260 Most Mentioned<\/div>\n<strong>Benson &amp; Bingham<\/strong> leads total mentions (880) but ranks #1 only 16% of the time. <strong>Naqvi Injury Law<\/strong> has fewer total mentions (808) but ranks #1 in 25% of its appearances \u2014 the highest #1 rate of any top firm. <strong>Eglet Law<\/strong> is even more dramatic: 416 mentions but #1 a whopping 24% of the time. Being mentioned often and being recommended first are two different games.\n<\/div>\n\n<div class=\"callout\">\n<strong>Share of Voice:<\/strong> Naqvi 7.2% \u00b7 Benson &amp; Bingham 6.8% \u00b7 Richard Harris 5.4% \u00b7 Eglet 3.7% \u00b7 Claggett &amp; Sykes 2.8%. These 5 firms control 25.9% of all AI recommendations. The remaining 1,239 firms share 74.1%.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 2: ECOSYSTEM OVERLAP ===== -->\n<section>\n<div class=\"tag\">Finding 02 \u00b7 Cross-Model Overlap<\/div>\n<h2>72.8% of Firms Are Known by Only One Ecosystem<\/h2>\n<p class=\"section-sub\">We grouped all channels into 3 ecosystems (OpenAI, Google, Perplexity) and checked: how many know each firm?<\/p>\n\n<div class=\"donut-wrap\">\n<svg width=\"210\" height=\"210\" viewBox=\"0 0 220 220\">\n<circle cx=\"110\" cy=\"110\" r=\"80\" fill=\"none\" stroke=\"#ef4444\" stroke-width=\"28\" stroke-dasharray=\"459.5 502.4\" stroke-dashoffset=\"125.6\" transform=\"rotate(-90 110 110)\"\/>\n<circle cx=\"110\" cy=\"110\" r=\"80\" fill=\"none\" stroke=\"#f59e0b\" stroke-width=\"28\" stroke-dasharray=\"80 502.4\" stroke-dashoffset=\"-333.9\" transform=\"rotate(-90 110 110)\"\/>\n<circle cx=\"110\" cy=\"110\" r=\"80\" fill=\"none\" stroke=\"#22c55e\" stroke-width=\"28\" stroke-dasharray=\"56.3 502.4\" stroke-dashoffset=\"-413.9\" transform=\"rotate(-90 110 110)\"\/>\n<text x=\"110\" y=\"105\" text-anchor=\"middle\" font-family=\"var(--theme-font-family)\" font-size=\"26\" font-weight=\"700\" fill=\"#fff\">1,267<\/text>\n<text x=\"110\" y=\"126\" text-anchor=\"middle\" font-family=\"var(--theme-font-family)\" font-size=\"11\" fill=\"#8888a0\">total firms<\/text>\n<\/svg>\n<div class=\"legend\">\n<div class=\"legend-item\"><div class=\"legend-dot\" style=\"background:#22c55e\"><\/div><div><div class=\"legend-pct\">11.2% \u2014 142 firms<\/div><div class=\"legend-label\">Known by all 3 ecosystems (the AI canon)<\/div><\/div><\/div>\n<div class=\"legend-item\"><div class=\"legend-dot\" style=\"background:#f59e0b\"><\/div><div><div class=\"legend-pct\">15.9% \u2014 202 firms<\/div><div class=\"legend-label\">Known by 2 ecosystems<\/div><\/div><\/div>\n<div class=\"legend-item\"><div class=\"legend-dot\" style=\"background:#ef4444\"><\/div><div><div class=\"legend-pct\">72.8% \u2014 923 firms<\/div><div class=\"legend-label\">Known by only 1 ecosystem<\/div><\/div><\/div>\n<\/div>\n<\/div>\n\n<div id=\"ecosystemChart\" style=\"margin:24px auto;text-align:center\"><\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 Pairwise Ecosystem Overlap<\/div>\nGoogle &amp; Perplexity share the most firms (27.3% Jaccard overlap), while OpenAI &amp; Google share only 19.4%. This suggests Perplexity and Google draw from more similar source material than OpenAI does. If your firm is visible in Perplexity, there's a decent chance Google knows you too \u2014 but ChatGPT may have never heard of you.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 3: DIRECT vs API ===== -->\n<section>\n<div class=\"tag\">Finding 03 \u00b7 User Interface vs API \u2014 The Headline Finding<\/div>\n<h2>Same Model, Different Answers: User Interface vs API Overlap Is 3-12%<\/h2>\n<p class=\"section-sub\">We tested the exact same prompts on ChatGPT, Gemini, and Perplexity via their web interface (Direct) and their API. The per-prompt overlap is shockingly low.<\/p>\n\n<div class=\"overlap-grid\" id=\"directApiGrid\"><\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 ChatGPT User Interface vs ChatGPT API for the same prompt<\/div>\nFor prompt \"Best personal injury lawyer in Las Vegas\" \u2014 <strong>ChatGPT User Interface<\/strong> (Web Search ON) recommended: Richard Harris, Sam &amp; Ash, Benson &amp; Bingham, and Hurtado Law. <strong>ChatGPT API<\/strong> (Web Search ON) for the same prompt returned: Morgan &amp; Morgan, Adam S. Kutner, Vegas Valley Injury Law, and The Accident Guys. <strong>Zero overlap.<\/strong> Different system prompts, different browsing behavior, entirely different recommendations.\n<\/div>\n\n<div class=\"callout\">\n<strong>Why this happens:<\/strong> The web UI has hidden system prompts, safety wrappers, and auto-browsing behavior the API doesn't have. ChatGPT's UI often triggers Bing browsing automatically; the API only searches if you explicitly pass the tool. These \"invisible layers\" dramatically change who gets recommended.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 4: SEARCH ON vs OFF ===== -->\n<section>\n<div class=\"tag\">Finding 04 \u00b7 Web Search ON vs OFF Delta<\/div>\n<h2>Toggle Web Search and 58-87% of Firms Change<\/h2>\n<p class=\"section-sub\">Every channel was tested with web search ON (live internet) and OFF (training data only). The overlap between the two modes:<\/p>\n\n<div class=\"bar-chart\" id=\"searchChart\"><\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 Gemini User Interface<\/div>\nGemini User Interface with web search OFF surfaces 113 unique firms. Turn search ON and it surfaces 204 firms \u2014 nearly double. But only 64 firms appear in both modes (25.3% overlap). That means <strong>140 firms only exist in one mode or the other.<\/strong> The training-data version recommends a completely different competitive landscape than the search-grounded version.\n<\/div>\n\n<div class=\"callout\">\n<strong>Perplexity is the most stable<\/strong> at 42.2% overlap (User Interface) and 35.6% (API) \u2014 because it always grounds in search. ChatGPT API is the least stable: just 12.5% overlap between search ON and OFF. If you're only tracking one mode, you're missing up to 87% of the picture.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 5: GOOGLE 4-WAY ===== -->\n<section>\n<div class=\"tag\">Finding 05 \u00b7 The Google 4-Way \u2014 The Money Slide<\/div>\n<h2>Same Google Brain, 4 Products, 4 Different Answers<\/h2>\n<p class=\"section-sub\">We tested the Google ecosystem across 4 product surfaces. Same underlying model family \u2014 radically different outputs.<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Google Product<\/th><th class=\"n\">Unique Firms<\/th><th class=\"n\">Avg\/Response<\/th><th>Channel Type<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Google AI Mode<\/td><td class=\"n\">290<\/td><td class=\"n\">12.0<\/td><td>Google Search AI<\/td><\/tr>\n<tr><td>Gemini User Interface (Web:ON)<\/td><td class=\"n\">204<\/td><td class=\"n\">8.2<\/td><td>User Interface<\/td><\/tr>\n<tr><td>Gemini API (Web:ON)<\/td><td class=\"n\">246<\/td><td class=\"n\">7.3<\/td><td>API<\/td><\/tr>\n<tr><td>Gemini API (Web:OFF)<\/td><td class=\"n\">156<\/td><td class=\"n\">7.6<\/td><td>API<\/td><\/tr>\n<tr><td>Gemini User Interface (Web:OFF)<\/td><td class=\"n\">113<\/td><td class=\"n\">5.7<\/td><td>User Interface<\/td><\/tr>\n<tr><td>Google Local Pack (SERP)<\/td><td class=\"n\">44<\/td><td class=\"n\">3.0<\/td><td>Traditional SERP<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 Naqvi Injury Law Across Google Products<\/div>\n<strong>Naqvi Injury Law<\/strong> in Google AI Mode: 137 mentions (avg rank 5.4). In Gemini User Interface: significantly fewer. In Local Pack: 52 appearances (rank #1 most often). The firm dominates differently in each Google product. A firm tracking only Local Pack rankings would miss that they're also the #1 most-mentioned firm in AI Mode.\n<\/div>\n\n<div class=\"callout\">\n<strong>AI Mode surfaces 6.6\u00d7 more firms than Local Pack<\/strong> (290 vs 44). The overlap between AI Mode and Gemini User Interface (Web:ON) is only 30%. If you're only tracking one Google channel, you're flying blind. This is the strongest argument for multi-channel AI visibility monitoring.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 6: EACH AI HAS A PERSONALITY ===== -->\n<section>\n<div class=\"tag\">Finding 06 \u00b7 AI Personality \u2014 Rank Disagreements<\/div>\n<h2>Each AI Ecosystem Ranks Firms Differently<\/h2>\n<p class=\"section-sub\">The same top firms get dramatically different average rank positions depending on which ecosystem you ask.<\/p>\n\n<div id=\"rankDotChart\" style=\"margin:24px auto;text-align:center\"><\/div>\n\n<table class=\"dt\">\n<thead><tr><th>Firm<\/th><th class=\"n\">OpenAI Avg Rank<\/th><th class=\"n\">Google Avg Rank<\/th><th class=\"n\">AI Mode Avg Rank<\/th><th class=\"n\">Perplexity Avg Rank<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Naqvi Injury Law<\/td><td class=\"n\">3.3<\/td><td class=\"n\">4.4<\/td><td class=\"n\">5.4<\/td><td class=\"n\">4.0<\/td><\/tr>\n<tr><td>Benson &amp; Bingham<\/td><td class=\"n\">4.9<\/td><td class=\"n\">4.1<\/td><td class=\"n\">5.8<\/td><td class=\"n\">5.1<\/td><\/tr>\n<tr><td>Richard Harris Law Firm<\/td><td class=\"n\">3.6<\/td><td class=\"n\">4.8<\/td><td class=\"n\">5.5<\/td><td class=\"n\">4.5<\/td><\/tr>\n<tr><td>Eglet Law<\/td><td class=\"n\">3.0<\/td><td class=\"n\">5.2<\/td><td class=\"n\">5.9<\/td><td class=\"n\">3.8<\/td><\/tr>\n<tr><td>Adam S. Kutner<\/td><td class=\"n\">4.2<\/td><td class=\"n\">5.5<\/td><td class=\"n\">7.7<\/td><td class=\"n\">5.1<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 Eglet Law<\/div>\n<strong>Eglet Law<\/strong> averages rank #3.0 in OpenAI (nearly always in the top 3) but drops to #5.9 in Google AI Mode. ChatGPT clearly \"likes\" Eglet more than Google does. Meanwhile, <strong>Richard Harris<\/strong> is consistently ranked #3-5 across all ecosystems \u2014 the most stable brand positioning of any top firm.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 7: PROMPT DIMENSION MAPPING ===== -->\n<section>\n<div class=\"tag\">Finding 07 \u00b7 Prompt Dimension Mapping<\/div>\n<h2>Which Query Types Trigger Which Firms?<\/h2>\n<p class=\"section-sub\">We tested 124 prompts across 8 dimensions. Some firms dominate generic queries; others only surface for specific injury types or trust-based queries.<\/p>\n\n<table class=\"dt\">\n<thead>\n<tr>\n  <th>Dimension<\/th>\n  <th class=\"n\">Prompts<\/th>\n  <th class=\"n\">Total Mentions<\/th>\n  <th>Top Firm<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n  <td>Generic \/ Broad<\/td>\n  <td class=\"n\">15<\/td>\n  <td class=\"n\">1,396<\/td>\n  <td>Naqvi Injury Law (151)<\/td>\n<\/tr>\n<tr>\n  <td>Injury Type<\/td>\n  <td class=\"n\">30<\/td>\n  <td class=\"n\">2,501<\/td>\n  <td>Benson &amp; Bingham (225)<\/td>\n<\/tr>\n<tr>\n  <td>Trust &amp; Credibility<\/td>\n  <td class=\"n\">15<\/td>\n  <td class=\"n\">1,376<\/td>\n  <td>Benson &amp; Bingham (123)<\/td>\n<\/tr>\n<tr>\n  <td>Situational \/ Scenario<\/td>\n  <td class=\"n\">20<\/td>\n  <td class=\"n\">1,592<\/td>\n  <td>Benson &amp; Bingham (140)<\/td>\n<\/tr>\n<tr>\n  <td>Location Specific<\/td>\n  <td class=\"n\">14<\/td>\n  <td class=\"n\">1,293<\/td>\n  <td>Naqvi Injury Law (110)<\/td>\n<\/tr>\n<tr>\n  <td>Constraints &amp; Practical<\/td>\n  <td class=\"n\">12<\/td>\n  <td class=\"n\">981<\/td>\n  <td>Naqvi Injury Law (69)<\/td>\n<\/tr>\n<tr>\n  <td>Comparison \/ Versus<\/td>\n  <td class=\"n\">10<\/td>\n  <td class=\"n\">885<\/td>\n  <td>Richard Harris (56)<\/td>\n<\/tr>\n<tr>\n  <td>Vegas-Specific<\/td>\n  <td class=\"n\">8<\/td>\n  <td class=\"n\">725<\/td>\n  <td>Richard Harris (53)<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"callout callout-green\">\n<strong>Niche Specialists:<\/strong> Some firms dramatically over-index on specific dimensions. <strong>Eglet Law<\/strong> gets 26% of its mentions from Trust &amp; Credibility queries (vs 12% expected) \u2014 it's the \"awards &amp; settlements\" firm. <strong>Paul Padda Law<\/strong> gets 45% from Injury Type queries (vs 24% expected) \u2014 AI sees it as the specialist. <strong>Cameron Law<\/strong> over-indexes on Constraints queries \u2014 the \"Spanish-speaking, 24-hour\" firm in AI's mind.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 8: GBP CORRELATION ===== -->\n<section>\n<div class=\"tag\">Finding 08 \u00b7 Fame Beats Quality<\/div>\n<h2>A 5-Star Firm Can Be Invisible to AI<\/h2>\n<p class=\"section-sub\">We checked whether Google ratings or review volume predict AI visibility. The answer: being great doesn't make you famous to AI.<\/p>\n\n<div id=\"fameChart\" style=\"margin:24px auto;text-align:center\"><\/div>\n\n<div class=\"prose\">\n<p>We compared every firm's <span class=\"hl\">Google star rating<\/span> against how often AI recommends them. The result? <strong>No connection at all.<\/strong> Higher-rated firms are not more likely to be recommended \u2014 if anything, they're slightly less likely. A perfect 5.0-star firm can be completely invisible to AI.<\/p>\n<p><span class=\"hl-g\">Review count<\/span> helps a little \u2014 firms with more reviews tend to get mentioned slightly more \u2014 but it's a weak signal. Having 800 reviews doesn't guarantee AI knows you exist.<\/p>\n<\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 This tells the whole story<\/div>\n<strong>PT Law:<\/strong> Perfect 5.0 rating, 932 reviews \u2014 but only 3 AI mentions across all channels.<br>\n<strong>Ladah Law Firm:<\/strong> 4.8 rating, 498 reviews, just 2 Local Pack appearances \u2014 but 268 AI mentions and 317 citations (3rd most-cited website).<br>\n<strong>Ladah has nearly 90\u00d7 more AI visibility than PT Law despite having fewer reviews and a lower star rating.<\/strong> The difference? Ladah's website content gets cited by AI. PT Law's doesn't. Being talked about online matters more than being liked.\n<\/div>\n<div class=\"callout callout-purple\">\n<strong>What AI actually reads:<\/strong> LLMs don't check your star rating. They absorb the volume of discussion about your business \u2014 blog posts, news articles, legal directory features, and content marketing. That's what enters training data and search results. Client satisfaction is invisible to AI unless it generates online content.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 9: THE INVISIBLE FIRM ===== -->\n<section>\n<div class=\"tag\">Finding 09 \u00b7 The Invisible Firm<\/div>\n<h2>5.0 Stars. 932 Reviews. Only 3 AI Mentions.<\/h2>\n<p class=\"section-sub\">We found firms with strong real-world presence \u2014 perfect ratings, hundreds of reviews \u2014 that AI barely knows exist.<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Firm<\/th><th class=\"n\">Rating<\/th><th class=\"n\">Reviews<\/th><th class=\"n\">Local Pack<\/th><th class=\"n\">AI Mentions<\/th><th>Verdict<\/th><\/tr><\/thead>\n<tbody>\n<tr><td><strong>PT Law<\/strong><\/td><td class=\"n\">5.0<\/td><td class=\"n\">932<\/td><td class=\"n\">2<\/td><td class=\"n\">3<\/td><td style=\"color:var(--red)\">AI Invisible<\/td><\/tr>\n<tr><td><strong>CVBN LAW<\/strong><\/td><td class=\"n\">4.9<\/td><td class=\"n\">504<\/td><td class=\"n\">2<\/td><td class=\"n\">5<\/td><td style=\"color:var(--red)\">AI Invisible<\/td><\/tr>\n<tr><td><strong>Ace Lakhani Law Firm<\/strong><\/td><td class=\"n\">5.0<\/td><td class=\"n\">468<\/td><td class=\"n\">1<\/td><td class=\"n\">6<\/td><td style=\"color:var(--red)\">AI Invisible<\/td><\/tr>\n<tr><td><strong>Hale Injury Law<\/strong><\/td><td class=\"n\">4.9<\/td><td class=\"n\">599<\/td><td class=\"n\">2<\/td><td class=\"n\">12<\/td><td style=\"color:var(--accent)\">Barely Visible<\/td><\/tr>\n<tr><td><strong>Tingey Injury Law Firm<\/strong><\/td><td class=\"n\">4.9<\/td><td class=\"n\">460<\/td><td class=\"n\">10<\/td><td class=\"n\">13<\/td><td style=\"color:var(--accent)\">Barely Visible<\/td><\/tr>\n<tr><td><strong>Maier Gutierrez Injury Lawyers<\/strong><\/td><td class=\"n\">5.0<\/td><td class=\"n\">574<\/td><td class=\"n\">12<\/td><td class=\"n\">18<\/td><td style=\"color:var(--accent)\">Barely Visible<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"callout\">\n<strong>PT Law<\/strong> is the most dramatic case: perfect 5.0 star rating, 932 reviews \u2014 nearly a thousand happy clients. But across 1,860 AI queries, it was mentioned only 3 times. <strong>Ladah Law Firm<\/strong> with fewer reviews (498) and a lower rating (4.8) has 268 AI mentions \u2014 nearly 90\u00d7 more AI visibility. The gap between real-world quality and AI awareness is massive, but solvable.\n<\/div>\n\n<div class=\"prose\">\n<p>These \"invisible firms\" represent the immediate opportunity for AI visibility optimization. All 6 firms have 4.9-5.0 star ratings and 460-932 reviews \u2014 strong fundamentals. But they're missing the content layer that LLMs consume: blog mentions, legal directory features, listicle placements, and third-party coverage that enters training data and search results.<\/p>\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 10: LOCAL PACK vs AI ===== -->\n<section>\n<div class=\"tag\">Finding 10 \u00b7 Local Pack \u2260 AI Recommendations<\/div>\n<h2>Different Channels, Different Winners<\/h2>\n<p class=\"section-sub\">93% of Local Pack firms appear somewhere in AI channels \u2014 but only 32% of top AI firms appear in the Local Pack. They're different games.<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Firm<\/th><th class=\"n\">Local Pack<\/th><th class=\"n\">AI Total<\/th><th class=\"n\">Google AI Mode<\/th><th>Pattern<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Naqvi Injury Law<\/td><td class=\"n\">52<\/td><td class=\"n\">808<\/td><td class=\"n\">137<\/td><td style=\"color:var(--green)\">Dominant Everywhere<\/td><\/tr>\n<tr><td>Jack Bernstein Injury Lawyers<\/td><td class=\"n\">0<\/td><td class=\"n\">202<\/td><td class=\"n\">25<\/td><td style=\"color:var(--accent3)\">AI Only \u2014 LP Invisible<\/td><\/tr>\n<tr><td>Gina Corena &amp; Associates<\/td><td class=\"n\">29<\/td><td class=\"n\">16<\/td><td class=\"n\">11<\/td><td style=\"color:var(--red)\">LP Only \u2014 AI Invisible<\/td><\/tr>\n<tr><td>Claggett &amp; Sykes<\/td><td class=\"n\">4<\/td><td class=\"n\">417<\/td><td class=\"n\">62<\/td><td style=\"color:var(--accent3)\">AI Dominant<\/td><\/tr>\n<tr><td>Dimopoulos Injury Law<\/td><td class=\"n\">15<\/td><td class=\"n\">86<\/td><td class=\"n\">19<\/td><td style=\"color:var(--accent4)\">Balanced<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"callout callout-green\">\n<strong>Jack Bernstein Injury Lawyers<\/strong> has zero Local Pack appearances but 202 AI mentions across 11 channels. This firm doesn't exist in traditional Google Search but lives in the AI recommendation canon. <strong>The reverse of an invisible firm \u2014 a firm that only exists in AI.<\/strong>\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 11: MENTIONS PER RESPONSE ===== -->\n<section>\n<div class=\"tag\">Finding 11 \u00b7 Discovery Gap Between Models<\/div>\n<h2>Google AI Mode Surfaces 2\u00d7 More Firms Than ChatGPT API<\/h2>\n<p class=\"section-sub\">Some channels give curated short lists; others produce exhaustive recommendations. If you're a smaller firm, the verbose channels are your best shot.<\/p>\n\n<div class=\"bar-chart\" id=\"avgChart\"><\/div>\n\n<div class=\"callout\">\n<strong>Google AI Mode averages 12.0 firms per response<\/strong> \u2014 the most generous recommender. ChatGPT API (Web Search ON) gives only 5.0. For lesser-known firms, AI Mode casts 2.4\u00d7 the net. Google products are exhaustive listers; ChatGPT is a curated recommender.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 12: CHANNEL EXCLUSIVES ===== -->\n<section>\n<div class=\"tag\">Finding 12 \u00b7 Channel-Exclusive Firms<\/div>\n<h2>148 Firms Exist Only in ChatGPT User Interface (Web Search ON)<\/h2>\n<p class=\"section-sub\">Each channel surfaces firms that no other channel mentions. These are \"platform-exclusive\" visibility opportunities.<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Channel<\/th><th class=\"n\">Exclusive Firms<\/th><th>What This Means<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>ChatGPT User Interface (Web:ON)<\/td><td class=\"n\">148<\/td><td>Browsing behavior surfaces unique results<\/td><\/tr>\n<tr><td>ChatGPT API (Web:ON)<\/td><td class=\"n\">110<\/td><td>API web search finds different pages than UI<\/td><\/tr>\n<tr><td>ChatGPT User Interface (Web:OFF)<\/td><td class=\"n\">54<\/td><td>Training data has firms API doesn't surface<\/td><\/tr>\n<tr><td>ChatGPT API (Web:OFF)<\/td><td class=\"n\">81<\/td><td>Different model version\/system prompt<\/td><\/tr>\n<tr><td>Gemini API (Web:ON)<\/td><td class=\"n\">77<\/td><td>Google grounding finds different firms<\/td><\/tr>\n<tr><td>Google AI Mode<\/td><td class=\"n\">76<\/td><td>Query fan-out discovers niche firms<\/td><\/tr>\n<tr><td>Perplexity API (Web:OFF)<\/td><td class=\"n\">62<\/td><td>Perplexity's training data is distinctive<\/td><\/tr>\n<tr><td>Gemini API (Web:OFF)<\/td><td class=\"n\">60<\/td><td>Gemini's base knowledge differs<\/td><\/tr>\n<\/tbody>\n<\/table>\n<\/section>\n\n<!-- ===== FINDING 13: CITATION SOURCE TRACKING ===== -->\n<section>\n<div class=\"tag\">Finding 13 \u00b7 Citation Source Tracking<\/div>\n<h2>Where Do AIs Get Their Information?<\/h2>\n<p class=\"section-sub\">We extracted 6,416 citation URLs from the User Interface channels. Law firm websites dominate (48.4%), followed by legal directories (13.1%) and review platforms (6.4%).<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Source Category<\/th><th class=\"n\">Citations<\/th><th class=\"n\">Share<\/th><th>Top Examples<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Law Firm Websites<\/td><td class=\"n\">3,112<\/td><td class=\"n\">48.4%<\/td><td>bensonbingham.com, ladahlaw.com, egletlaw.com, naqvilaw.com<\/td><\/tr>\n<tr><td>Other (News, Reddit, Forbes)<\/td><td class=\"n\">2,047<\/td><td class=\"n\">31.9%<\/td><td>forbes.com, reddit.com, cameronlawlv.com<\/td><\/tr>\n<tr><td>Legal Directories<\/td><td class=\"n\">841<\/td><td class=\"n\">13.1%<\/td><td>superlawyers.com, justia.com, avvo.com<\/td><\/tr>\n<tr><td>Review Platforms<\/td><td class=\"n\">413<\/td><td class=\"n\">6.4%<\/td><td>yelp.com, bbb.org<\/td><\/tr>\n<tr><td>News \/ Media<\/td><td class=\"n\">13<\/td><td class=\"n\">0.2%<\/td><td>reviewjournal.com, lasvegassun.com<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 Top 5 Most-Cited Domains<\/div>\n<strong>bensonbingham.com<\/strong> (456 citations) \u2014 the most cited law firm website across all channels. <strong>superlawyers.com<\/strong> (342) \u2014 the #1 legal directory source. <strong>ladahlaw.com<\/strong> (317) \u2014 surprisingly high given Ladah's modest Local Pack presence (2 appearances). <strong>yelp.com<\/strong> (243) \u2014 Google AI Mode cites Yelp heavily (186 of those 243). <strong>naqvilaw.com<\/strong> (230) \u2014 strong firm site, but behind Benson &amp; Bingham and Ladah in citations despite leading in AI mentions.\n<\/div>\n\n<div class=\"callout callout-green\">\n<strong>The Ladah Law Firm paradox:<\/strong> Ladah has only 2 Local Pack appearances and 498 reviews, but 212 AI mentions and 317 citations \u2014 the 3rd most-cited law firm website. Their website is clearly optimized for the content that AI scrapes. This is the strongest evidence that <strong>website content strategy drives AI visibility<\/strong> independently from traditional SEO.\n<\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 ChatGPT's Sources vs Google AI Mode<\/div>\n<strong>ChatGPT User Interface<\/strong> cites forbes.com more than any other domain (70 citations) \u2014 it trusts mainstream publications. <strong>Google AI Mode<\/strong> cites yelp.com heavily (213 citations) \u2014 it pulls from its own ecosystem. <strong>Perplexity<\/strong> cites ladahlaw.com most (86 citations) \u2014 it goes directly to firm websites. Each AI has a different \"source personality.\"\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 15: AI PERSONALITY \/ SENTIMENT ===== -->\n<section>\n<div class=\"tag\">Finding 14 \u00b7 Each AI Has a Voice<\/div>\n<h2>Same Firm, Different Story \u2014 How Each AI Describes the Top Firms<\/h2>\n<p class=\"section-sub\">Beyond just recommending different firms, each AI model frames the same firm with a different personality and emphasis.<\/p>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 How models describe Naqvi Injury Law for \"Best PI lawyer in Las Vegas\"<\/div>\n<strong>Gemini User Interface:<\/strong> \"11-time 'Best of Las Vegas' Gold Winner; extremely high client satisfaction\" \u2014 <em>credential-first, award-focused<\/em><br>\n<strong>Google AI Mode:<\/strong> Shows Google Maps card with rating (4.9), address, and image \u2014 <em>structured data, visual, local-pack style<\/em><br>\n<strong>ChatGPT User Interface:<\/strong> Ranked in a numbered list with brief qualifier \u2014 <em>concise, editorial tone<\/em>\n<\/div>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 How Gemini frames the top 4 firms (structured table format)<\/div>\n<strong>Naqvi:<\/strong> \"11-time Best of Las Vegas Gold Winner; extremely high client satisfaction\" \u2014 <em>awards + client sentiment<\/em><br>\n<strong>Eglet Adams:<\/strong> \"Robert Eglet named Trial Lawyer of the Year (2026); known for billion-dollar verdicts\" \u2014 <em>individual lawyer + headline verdicts<\/em><br>\n<strong>Benson &amp; Bingham:<\/strong> \"25+ years experience; record-setting settlements ($30M+)\" \u2014 <em>experience + settlement size<\/em><br>\n<strong>Richard Harris:<\/strong> \"Over 40 years in Nevada; one of the largest and most established firms\" \u2014 <em>longevity + scale<\/em>\n<\/div>\n\n<div class=\"callout callout-purple\">\n<strong>Gemini leads with credentials and data.<\/strong> ChatGPT leads with editorial recommendations. Google AI Mode leads with structured local data (Maps cards, ratings). Perplexity leads with citations and source attribution. A firm's AI optimization strategy should match how each model frames businesses \u2014 Gemini wants awards, ChatGPT wants narrative, Google wants GBP signals.\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 16: STABILITY ===== -->\n<section>\n<div class=\"tag\">Finding 15 \u00b7 Stability \/ Reproducibility<\/div>\n<h2>Ask Twice, Get Different Answers: Only 23.8% Overlap Between Runs<\/h2>\n<p class=\"section-sub\">We ran 20 prompts 4 times each across all 13 channels \u2014 1,040 queries \u2014 to test reproducibility. The results are sobering.<\/p>\n\n<div id=\"stabilityChart\" style=\"margin:24px auto;text-align:center\"><\/div>\n\n<table class=\"dt\">\n<thead><tr><th>Channel<\/th><th class=\"n\">Set Overlap<\/th><th class=\"n\">Core %<\/th><th class=\"n\">Stochastic %<\/th><th>Stability<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Gemini User Interface (Web:OFF)<\/td><td class=\"n\">38.8%<\/td><td class=\"n\">28.4%<\/td><td class=\"n\">71.6%<\/td><td style=\"color:var(--green)\">Most Stable<\/td><\/tr>\n<tr><td>Perplexity API (Web:OFF)<\/td><td class=\"n\">31.9%<\/td><td class=\"n\">24.0%<\/td><td class=\"n\">76.0%<\/td><td style=\"color:var(--green)\">Stable<\/td><\/tr>\n<tr><td>Perplexity User Interface (Web:ON)<\/td><td class=\"n\">28.7%<\/td><td class=\"n\">23.0%<\/td><td class=\"n\">77.0%<\/td><td style=\"color:var(--accent4)\">Moderate<\/td><\/tr>\n<tr><td>Perplexity User Interface (Web:OFF)<\/td><td class=\"n\">28.4%<\/td><td class=\"n\">23.6%<\/td><td class=\"n\">76.4%<\/td><td style=\"color:var(--accent4)\">Moderate<\/td><\/tr>\n<tr><td>ChatGPT User Interface (Web:OFF)<\/td><td class=\"n\">24.7%<\/td><td class=\"n\">18.8%<\/td><td class=\"n\">81.2%<\/td><td style=\"color:var(--accent4)\">Moderate<\/td><\/tr>\n<tr><td>Google AI Mode<\/td><td class=\"n\">22.3%<\/td><td class=\"n\">15.6%<\/td><td class=\"n\">84.4%<\/td><td style=\"color:var(--accent4)\">Moderate<\/td><\/tr>\n<tr><td>ChatGPT User Interface (Web:ON)<\/td><td class=\"n\">14.3%<\/td><td class=\"n\">7.6%<\/td><td class=\"n\">92.4%<\/td><td style=\"color:var(--red)\">Volatile<\/td><\/tr>\n<tr><td>ChatGPT API (Web:OFF)<\/td><td class=\"n\">8.8%<\/td><td class=\"n\">4.9%<\/td><td class=\"n\">95.1%<\/td><td style=\"color:var(--red)\">Very Volatile<\/td><\/tr>\n<tr><td>ChatGPT API (Web:ON)<\/td><td class=\"n\">6.5%<\/td><td class=\"n\">1.5%<\/td><td class=\"n\">98.5%<\/td><td style=\"color:var(--red)\">Extremely Volatile<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 ChatGPT API (Web:ON)<\/div>\nOnly <strong>6.5% set overlap<\/strong> between repeat runs. 98.5% of recommendations are stochastic \u2014 appearing in 2 or fewer of 4 runs. Ask ChatGPT's API the same question 4 times and you'll get almost entirely different law firms each time. Only 1.5% of recommendations are \"core\" (appear in 3+ runs).\n<\/div>\n\n<div class=\"callout\">\n<strong>82.3% of all AI recommendations are stochastic.<\/strong> They appear randomly \u2014 in 2 or fewer of 4 runs. Only 17.7% are \"core\" recommendations that show up reliably. Any AI visibility study that queries each model only once is measuring signal mixed with substantial noise. You need 3-5 runs to separate core from stochastic.\n<\/div>\n\n<div class=\"prose\">\n<p><span class=\"hl-g\">User Interface channels are more stable than API<\/span> (26.0% vs 21.7% overlap). <span class=\"hl-p\">Perplexity is the most stable ecosystem<\/span> (28.6% average overlap), while <span class=\"hl\">OpenAI is the least stable<\/span> (13.5%). Gemini User Interface (Web:OFF) is the single most stable channel at 38.8% \u2014 its training-data-only mode gives the most consistent answers.<\/p>\n<\/div>\n<\/section>\n\n<!-- ===== FINDING 17: KENDALL TAU ===== -->\n<section>\n<div class=\"tag\">Finding 16 \u00b7 Rank Consistency (Kendall's Tau)<\/div>\n<h2>Firms That Appear Keep Their Position \u2014 But Most Don't Appear<\/h2>\n<p class=\"section-sub\">Kendall's \u03c4 measures whether firms maintain the same rank ordering across repeat runs. Average: \u03c4 = 0.370 (moderate consistency).<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Channel<\/th><th class=\"n\">Kendall \u03c4<\/th><th class=\"n\">Set Overlap<\/th><th>Interpretation<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Perplexity User Interface (Web:OFF)<\/td><td class=\"n\">0.610<\/td><td class=\"n\">28.4%<\/td><td style=\"color:var(--green)\">Best rank consistency<\/td><\/tr>\n<tr><td>Perplexity User Interface (Web:ON)<\/td><td class=\"n\">0.590<\/td><td class=\"n\">28.7%<\/td><td style=\"color:var(--green)\">Strong rank consistency<\/td><\/tr>\n<tr><td>ChatGPT User Interface (Web:ON)<\/td><td class=\"n\">0.583<\/td><td class=\"n\">14.3%<\/td><td style=\"color:var(--accent4)\">Good ranks, volatile sets<\/td><\/tr>\n<tr><td>ChatGPT User Interface (Web:OFF)<\/td><td class=\"n\">0.558<\/td><td class=\"n\">24.7%<\/td><td style=\"color:var(--accent4)\">Good ranks, moderate sets<\/td><\/tr>\n<tr><td>Gemini User Interface (Web:OFF)<\/td><td class=\"n\">0.500<\/td><td class=\"n\">38.8%<\/td><td style=\"color:var(--accent4)\">Most stable overall<\/td><\/tr>\n<tr><td>Google AI Mode<\/td><td class=\"n\">0.221<\/td><td class=\"n\">22.3%<\/td><td style=\"color:var(--red)\">Weak rank consistency<\/td><\/tr>\n<tr><td>Gemini User Interface (Web:ON)<\/td><td class=\"n\">0.112<\/td><td class=\"n\">21.4%<\/td><td style=\"color:var(--red)\">Poor rank consistency<\/td><\/tr>\n<tr><td>ChatGPT API (Web:ON)<\/td><td class=\"n\">-0.289<\/td><td class=\"n\">6.5%<\/td><td style=\"color:var(--red)\">Inverted rankings<\/td><\/tr>\n<\/tbody>\n<\/table>\n\n<div class=\"example-box\">\n<div class=\"ex-label\">Real Example \u2014 The Stability Paradox<\/div>\n<strong>ChatGPT User Interface (Web:ON)<\/strong> has excellent rank consistency (\u03c4 = 0.583) but terrible set overlap (14.3%). It can't decide WHICH firms to mention \u2014 but when it does mention one, it puts it in roughly the same position. <strong>Gemini User Interface (Web:OFF)<\/strong> is the opposite: highest set overlap (38.8%) but moderate \u03c4 (0.500). It reliably mentions the same firms but shuffles their order.\n<\/div>\n\n<div class=\"callout callout-purple\">\n<strong>The practical implication:<\/strong> A firm like Naqvi Injury Law that appears as a \"core\" recommendation (3+ of 4 runs) in most channels has genuinely strong AI mindshare. A firm that appears once and disappears is riding noise, not signal. PromptLocal.ai tracks core vs stochastic visibility \u2014 because a single snapshot is meaningless.\n<\/div>\n<\/section>\n\n<!-- ===== LIMITATIONS ===== -->\n\n<!-- ===== SO WHAT ===== -->\n<section>\n<div class=\"tag\">So What? \u00b7 Implications for Law Firms<\/div>\n<h2>What Should a PI Lawyer in Las Vegas Actually Do?<\/h2>\n\n<div class=\"prose\">\n<p><strong>1. Track all channels, not just one.<\/strong> User Interface vs API gives different results. Search ON vs OFF gives different results. Google AI Mode vs Local Pack gives different results. A firm that's invisible in one channel may be dominant in another. Multi-channel monitoring is not optional.<\/p>\n\n<p><strong>2. Blog coverage and listicle placement matter more than star ratings.<\/strong> Firms like Ladah (212 AI mentions, 4.8 stars, 498 reviews) crush firms like Gina Corena (16 AI mentions, 5.0 stars, 836 reviews). The difference? Content footprint. Being talked about on legal blogs, featured in \"best of\" listicles, and covered by local press enters training data and live search results that AI consumes.<\/p>\n\n<p><strong>3. Google AI Mode is the new battleground.<\/strong> At 12 firms per response and 290 unique firms, it's the most generous and most important AI discovery channel for local businesses. It pulls from Google's full search index, Maps, and GBP data. Firms not optimizing for this channel are leaving AI visibility on the table.<\/p>\n\n<p><strong>4. The \"invisible firm\" opportunity is real.<\/strong> Maier Gutierrez has 574 reviews, 5.0 stars, 11 Local Pack appearances \u2014 and 2 AI mentions. The gap between real-world quality and AI awareness is a solvable problem. Content marketing, schema markup, and directory presence can move firms from invisible to visible within weeks.<\/p>\n\n<p><strong>5. Niche positioning works.<\/strong> Firms like Eglet Law (over-indexes on trust queries) and Paul Padda Law (over-indexes on injury-type queries) prove that you don't need to win every query \u2014 you need to own your niche across all channels.<\/p>\n<\/div>\n<\/section>\n\n<!-- ===== METHODOLOGY ===== -->\n<section>\n<div class=\"tag\">Methodology<\/div>\n<h2>How We Built This Study<\/h2>\n<div class=\"prose\">\n<p><strong>124 prompts<\/strong> across 8 dimensions \u2014 all recommendation-only, designed to guarantee business name mentions:<\/p>\n\n<table class=\"dt\">\n<thead><tr><th>Dimension<\/th><th class=\"n\">Prompts<\/th><th>Purpose<\/th><\/tr><\/thead>\n<tbody>\n<tr><td>Injury Type<\/td><td class=\"n\">30<\/td><td>Case-type visibility \u2014 car, truck, slip &amp; fall, med mal, rideshare, etc.<\/td><\/tr>\n<tr><td>Situational \/ Scenario<\/td><td class=\"n\">20<\/td><td>Narrative-based with explicit recommendation ask \u2014 real-world phrasing<\/td><\/tr>\n<tr><td>Generic \/ Broad<\/td><td class=\"n\">15<\/td><td>Baseline \u2014 who shows up for generic PI lawyer queries<\/td><\/tr>\n<tr><td>Trust &amp; Credibility<\/td><td class=\"n\">15<\/td><td>Reviews, settlements, awards \u2014 what trust signals does AI surface<\/td><\/tr>\n<tr><td>Location Specific<\/td><td class=\"n\">14<\/td><td>Neighborhood and venue-specific \u2014 tests local knowledge depth<\/td><\/tr>\n<tr><td>Constraints &amp; Practical<\/td><td class=\"n\">12<\/td><td>Fees, language, availability \u2014 filtered recommendation queries<\/td><\/tr>\n<tr><td>Comparison \/ Versus<\/td><td class=\"n\">10<\/td><td>Competitive rankings, firm-type filters \u2014 forces named comparisons<\/td><\/tr>\n<tr><td>Vegas-Specific<\/td><td class=\"n\">8<\/td><td>Casino, entertainment, pool club, convention injuries \u2014 unique to Las Vegas<\/td><\/tr>\n<tr style=\"font-weight:bold\"><td>TOTAL<\/td><td class=\"n\">124<\/td><td><\/td><\/tr>\n<\/tbody>\n<\/table>\n<p><strong>3 models:<\/strong> ChatGPT (GPT-5.4-mini), Gemini (2.5 Pro), Perplexity (Sonar).<\/p>\n<p><strong>3 channels:<\/strong> (1) User Interface \u2014 prompting AI interfaces directly through their consumer web products. (2) API \u2014 same models queried programmatically through their respective APIs. (3) Google Search \u2014 Local Pack + Organic results.<\/p>\n<p><strong>Entity resolution:<\/strong> 1,872 raw name variants collapsed through normalization, manual merge rules, and deduplication. Non-firm entities (directories, insurers, retailers) excluded. Final count: 1,267 unique law firms.<\/p>\n<p><strong>Full response analysis:<\/strong> Raw AI responses captured with full markdown text, citation URLs, links attached, and source metadata. 6,416 citation URLs extracted from User Interface channels across 363 unique domains.<\/p>\n<p><strong>Ground truth:<\/strong> Google Local Pack (89\/124 prompts triggered, 44 unique firms, with ratings\/reviews) and Organic results (124 prompts, 168 unique domains) as baseline comparison.<\/p>\n<\/div>\n<\/section>\n\n<div class=\"footer\">\n<p>Research by <a href=\"https:\/\/promptlocal.ai\">PromptLocal.ai<\/a> \u2014 AI Visibility Tracking for Local Businesses<\/p>\n<\/div>\n\n<\/div>\n\n<script>\n\/\/ === CHART HELPERS ===\nfunction svgEl(tag, attrs={}) {\n  const el = document.createElementNS('http:\/\/www.w3.org\/2000\/svg', tag);\n  for (const [k,v] of Object.entries(attrs)) el.setAttribute(k, v);\n  return el;\n}\n\n\/\/ === CHART 1: TOP 15 LEADERBOARD (horizontal bar) ===\nconst top15=[\n{n:\"Benson & Bingham\",m:880,r:4.6,r1:16,sov:\"6.8%\"},\n{n:\"Naqvi Injury Law\",m:808,r:4.0,r1:25,sov:\"7.2%\"},\n{n:\"Richard Harris Law Firm\",m:684,r:4.3,r1:17,sov:\"5.4%\"},\n{n:\"Claggett & Sykes\",m:417,r:4.9,r1:15,sov:\"2.8%\"},\n{n:\"Eglet Law\",m:416,r:4.3,r1:24,sov:\"3.7%\"},\n{n:\"Adam S. Kutner\",m:378,r:5.0,r1:12,sov:\"2.1%\"},\n{n:\"Ladah Law Firm\",m:222,r:4.7,r1:17,sov:\"2.0%\"},\n{n:\"Shook & Stone\",m:203,r:5.1,r1:12,sov:\"1.7%\"},\n{n:\"Jack Bernstein\",m:202,r:4.1,r1:22,sov:\"1.5%\"},\n{n:\"Battle Born Injury\",m:153,r:5.1,r1:8,sov:\"1.4%\"},\n{n:\"Lerner and Rowe\",m:145,r:3.8,r1:13,sov:\"1.1%\"},\n{n:\"Henness & Haight\",m:136,r:6.2,r1:7,sov:\"1.0%\"},\n{n:\"Paul Powell Law\",m:128,r:5.4,r1:13,sov:\"0.9%\"},\n{n:\"De Castroverde\",m:123,r:5.5,r1:14,sov:\"0.9%\"},\n{n:\"Cameron Law\",m:113,r:6.7,r1:5,sov:\"0.8%\"},\n];\nconst mx=top15[0].m;const c1=document.getElementById('top15chart');\nconst clrs=['#ff6b35','#ff8555','#ffa075','#ffb595','#ffcab5'];\ntop15.forEach((f,i)=>{const p=(f.m\/mx*100).toFixed(0);const c=i<5?clrs[i]:i<10?'#6366f1':'#4a4a6a';\nc1.innerHTML+=`<div class=\"bar-row\"><div class=\"bar-label\">${f.n}<\/div><div class=\"bar-track\"><div class=\"bar-fill\" style=\"width:${p}%;background:${c}\">${f.m}<\/div><\/div><div class=\"bar-meta\">rank ${f.r} \u00b7 #1 ${f.r1}%<\/div><\/div>`;});\n\n\/\/ === CHART 2: ECOSYSTEM OVERLAP  ===\n(function(){\nconst el = document.getElementById('ecosystemChart');\nif(!el) return;\nconst W=700, H=200, ml=10, mr=30;\nconst data = [{l:\"1 ecosystem\\n(longtail)\",c:923,p:72.8,col:\"#EF4444\"},{l:\"2 ecosystems\",c:202,p:15.9,col:\"#F59E0B\"},{l:\"All 3\\n(consensus)\",c:142,p:11.2,col:\"#22C55E\"}];\nconst total=data.reduce((s,d)=>s+d.c,0);\nconst bW=W-ml-mr;\nlet svg=`<svg viewBox=\"0 0 ${W} ${H}\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" style=\"width:100%;max-width:1000px\">`;\nsvg+=`<text x=\"${W\/2}\" y=\"18\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">How many ecosystems know each firm?<\/text>`;\nlet x=ml;\ndata.forEach(d=>{\n  const w=d.c\/total*bW;\n  svg+=`<rect x=\"${x}\" y=\"35\" width=\"${w}\" height=\"80\" fill=\"${d.col}\" rx=\"4\"\/>`;\n  svg+=`<text x=\"${x+w\/2}\" y=\"68\" text-anchor=\"middle\" fill=\"white\" font-size=\"18\" font-weight=\"700\" font-family=\"var(--theme-font-family)\">${d.c}<\/text>`;\n  svg+=`<text x=\"${x+w\/2}\" y=\"85\" text-anchor=\"middle\" fill=\"white\" font-size=\"11\" font-family=\"var(--theme-font-family)\" opacity=\"0.9\">(${d.p}%)<\/text>`;\n  const lines=d.l.split('\\n');\n  lines.forEach((line,li)=>{\n    svg+=`<text x=\"${x+w\/2}\" y=\"${130+li*14}\" text-anchor=\"middle\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${line}<\/text>`;\n  });\n  x+=w+4;\n});\nsvg+=`<text x=\"${W\/2}\" y=\"${H-8}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\" font-style=\"italic\">Only 11.2% of firms are in the universal AI canon. Nearly three-quarters are known by just one ecosystem.<\/text>`;\nsvg+=`<\/svg>`;\nel.innerHTML=svg;\n})();\n\n\/\/ === CHART 3: DIRECT vs API OVERLAP CARDS ===\nconst da=[\n{m:\"ChatGPT\",s:\"OFF\",ov:16.5,pp:4.7},{m:\"ChatGPT\",s:\"ON\",ov:16.0,pp:3.0},\n{m:\"Gemini\",s:\"OFF\",ov:30.0,pp:12.5},{m:\"Gemini\",s:\"ON\",ov:26.8,pp:7.6},\n{m:\"Perplexity\",s:\"OFF\",ov:28.0,pp:9.7},{m:\"Perplexity\",s:\"ON\",ov:27.3,pp:9.0},\n];\nconst dg=document.getElementById('directApiGrid');\nda.forEach(d=>{const c=d.pp<5?'#ef4444':d.pp<10?'#f59e0b':'#22c55e';\ndg.innerHTML+=`<div class=\"ov-card\"><h4>${d.m} \u00b7 Web Search ${d.s}<\/h4><div class=\"sub\">User Interface vs API firm overlap<\/div><div class=\"ov-bar\"><div class=\"ov-fill\" style=\"width:${d.ov}%;background:${c}\"><\/div><\/div><div style=\"display:flex;justify-content:space-between;align-items:baseline\"><div class=\"ov-pct\" style=\"color:${c}\">${d.pp}%<\/div><div style=\"font-size:11px;color:var(--dim)\">per-prompt overlap<\/div><\/div><\/div>`;});\n\n\/\/ === CHART 4: SEARCH ON vs OFF  ===\n(function(){\nconst el=document.getElementById('searchChart');\nif(!el) return;\nconst data=[\n  {l:\"ChatGPT User Interface\",both:63,on_only:262,off_only:77},\n  {l:\"ChatGPT API\",both:45,on_only:211,off_only:104},\n  {l:\"Gemini User Interface\",both:64,on_only:140,off_only:49},\n  {l:\"Gemini API\",both:54,on_only:192,off_only:102},\n  {l:\"Perplexity User Interface\",both:105,on_only:71,off_only:73},\n  {l:\"Perplexity API\",both:129,on_only:115,off_only:118},\n];\nconst W=700,H=260,ml=140,mr=50,barH=28,gap=8;\nlet svg=`<svg viewBox=\"0 0 ${W} ${H}\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" style=\"width:100%;max-width:1000px\">`;\nsvg+=`<text x=\"${W\/2}\" y=\"16\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Firm overlap: Web Search ON vs OFF<\/text>`;\nconst maxVal=Math.max(...data.map(d=>d.both+d.on_only+d.off_only));\nconst bW=W-ml-mr;\ndata.forEach((d,i)=>{\n  const y=30+i*(barH+gap);\n  const total=d.both+d.on_only+d.off_only;\n  const wBoth=d.both\/maxVal*bW;\n  const wOn=d.on_only\/maxVal*bW;\n  const wOff=d.off_only\/maxVal*bW;\n  svg+=`<text x=\"${ml-8}\" y=\"${y+barH\/2+4}\" text-anchor=\"end\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${d.l}<\/text>`;\n  \/\/ Off only (left)\n  svg+=`<rect x=\"${ml}\" y=\"${y}\" width=\"${wOff}\" height=\"${barH}\" fill=\"#EF4444\" rx=\"2\" opacity=\"0.7\"\/>`;\n  \/\/ Both (middle)\n  svg+=`<rect x=\"${ml+wOff}\" y=\"${y}\" width=\"${wBoth}\" height=\"${barH}\" fill=\"#22C55E\" rx=\"2\"\/>`;\n  \/\/ On only (right)\n  svg+=`<rect x=\"${ml+wOff+wBoth}\" y=\"${y}\" width=\"${wOn}\" height=\"${barH}\" fill=\"#6366F1\" rx=\"2\" opacity=\"0.7\"\/>`;\n  \/\/ Total label\n  const pctBoth=(d.both\/total*100).toFixed(0);\n  svg+=`<text x=\"${ml+wOff+wBoth+wOn+6}\" y=\"${y+barH\/2+4}\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${pctBoth}%<\/text>`;\n});\n\/\/ Legend\nconst ly=H-18;\nsvg+=`<rect x=\"${ml}\" y=\"${ly}\" width=\"12\" height=\"12\" fill=\"#EF4444\" opacity=\"0.7\" rx=\"2\"\/><text x=\"${ml+16}\" y=\"${ly+10}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Search OFF only<\/text>`;\nsvg+=`<rect x=\"${ml+120}\" y=\"${ly}\" width=\"12\" height=\"12\" fill=\"#22C55E\" rx=\"2\"\/><text x=\"${ml+136}\" y=\"${ly+10}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Both modes<\/text>`;\nsvg+=`<rect x=\"${ml+220}\" y=\"${ly}\" width=\"12\" height=\"12\" fill=\"#6366F1\" opacity=\"0.7\" rx=\"2\"\/><text x=\"${ml+236}\" y=\"${ly+10}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Search ON only<\/text>`;\nsvg+=`<\/svg>`;\nel.innerHTML=svg;\n})();\n\n\/\/ === CHART 5: RANK DISAGREEMENTS DOT PLOT  ===\n(function(){\nconst el=document.getElementById('rankDotChart');\nif(!el) return;\nconst firms=[\n  {n:\"Adam S. Kutner\",openai:4.2,google:5.5,aim:7.7,pplx:5.1},\n  {n:\"Eglet Law\",openai:3.0,google:5.2,aim:5.9,pplx:3.8},\n  {n:\"Shook & Stone\",openai:4.8,google:5.4,aim:7.2,pplx:4.6},\n  {n:\"Benson & Bingham\",openai:4.9,google:4.1,aim:5.8,pplx:5.1},\n  {n:\"Naqvi Injury Law\",openai:3.3,google:4.4,aim:5.4,pplx:4.0},\n  {n:\"Richard Harris\",openai:3.6,google:4.8,aim:5.5,pplx:4.5},\n  {n:\"Claggett & Sykes\",openai:5.1,google:4.5,aim:6.3,pplx:4.8},\n  {n:\"Ladah Law Firm\",openai:4.0,google:5.2,aim:5.8,pplx:4.5},\n];\n\/\/ Sort by max spread\nfirms.sort((a,b)=>{\n  const sa=Math.max(a.openai,a.google,a.aim,a.pplx)-Math.min(a.openai,a.google,a.aim,a.pplx);\n  const sb=Math.max(b.openai,b.google,b.aim,b.pplx)-Math.min(b.openai,b.google,b.aim,b.pplx);\n  return sb-sa;\n});\nconst W=700,ml=160,mr=60,dotR=6,rowH=32;\nconst H=40+firms.length*rowH+40;\nconst chartW=W-ml-mr;\nconst minR=1,maxR=10;\nconst scale=v=>(v-minR)\/(maxR-minR)*chartW;\n\nlet svg=`<svg viewBox=\"0 0 ${W} ${H}\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" style=\"width:100%;max-width:1000px\">`;\nsvg+=`<text x=\"${W\/2}\" y=\"16\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Biggest rank disagreements across ecosystems<\/text>`;\n\/\/ Axis\nfor(let r=1;r<=10;r++){\n  const x=ml+scale(r);\n  svg+=`<line x1=\"${x}\" y1=\"28\" x2=\"${x}\" y2=\"${30+firms.length*rowH}\" stroke=\"#2a2a3a\" stroke-width=\"0.5\"\/>`;\n  svg+=`<text x=\"${x}\" y=\"${30+firms.length*rowH+14}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${r}<\/text>`;\n}\nsvg+=`<text x=\"${ml+chartW\/2}\" y=\"${H-4}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Average rank position (lower = ranked higher)<\/text>`;\n\nconst colors={openai:\"#6366F1\",google:\"#F59E0B\",aim:\"#FF6B35\",pplx:\"#00D4AA\"};\nfirms.forEach((f,i)=>{\n  const y=36+i*rowH;\n  svg+=`<text x=\"${ml-8}\" y=\"${y+4}\" text-anchor=\"end\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${f.n}<\/text>`;\n  \/\/ Connection line\n  const vals=[f.openai,f.google,f.aim,f.pplx];\n  const minV=Math.min(...vals),maxV=Math.max(...vals);\n  svg+=`<line x1=\"${ml+scale(minV)}\" y1=\"${y}\" x2=\"${ml+scale(maxV)}\" y2=\"${y}\" stroke=\"#3a3a4a\" stroke-width=\"2\"\/>`;\n  \/\/ Dots\n  svg+=`<circle cx=\"${ml+scale(f.openai)}\" cy=\"${y}\" r=\"${dotR}\" fill=\"${colors.openai}\"\/>`;\n  svg+=`<circle cx=\"${ml+scale(f.google)}\" cy=\"${y}\" r=\"${dotR}\" fill=\"${colors.google}\"\/>`;\n  svg+=`<circle cx=\"${ml+scale(f.aim)}\" cy=\"${y}\" r=\"${dotR}\" fill=\"${colors.aim}\"\/>`;\n  svg+=`<circle cx=\"${ml+scale(f.pplx)}\" cy=\"${y}\" r=\"${dotR}\" fill=\"${colors.pplx}\"\/>`;\n});\n\/\/ Legend\nconst lx=ml;const ly=H-20;\nsvg+=`<circle cx=\"${lx}\" cy=\"${ly}\" r=\"5\" fill=\"${colors.openai}\"\/><text x=\"${lx+10}\" y=\"${ly+4}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">ChatGPT<\/text>`;\nsvg+=`<circle cx=\"${lx+80}\" cy=\"${ly}\" r=\"5\" fill=\"${colors.google}\"\/><text x=\"${lx+90}\" y=\"${ly+4}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Gemini<\/text>`;\nsvg+=`<circle cx=\"${lx+150}\" cy=\"${ly}\" r=\"5\" fill=\"${colors.aim}\"\/><text x=\"${lx+160}\" y=\"${ly+4}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">AI Mode<\/text>`;\nsvg+=`<circle cx=\"${lx+230}\" cy=\"${ly}\" r=\"5\" fill=\"${colors.pplx}\"\/><text x=\"${lx+240}\" y=\"${ly+4}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Perplexity<\/text>`;\nsvg+=`<\/svg>`;\nel.innerHTML=svg;\n})();\n\n\/\/ === CHART 6: AVG MENTIONS PER RESPONSE ===\nconst ad=[\n{l:\"Google AI Mode\",v:12.0,c:\"#ff6b35\"},{l:\"ChatGPT Interface (Web:ON)\",v:9.0,c:\"#6366f1\"},\n{l:\"Perplexity API (Web:OFF)\",v:8.8,c:\"#00d4aa\"},{l:\"Gemini Interface (Web:ON)\",v:8.2,c:\"#f59e0b\"},\n{l:\"Perplexity API (Web:ON)\",v:8.1,c:\"#00d4aa\"},{l:\"Gemini API (Web:OFF)\",v:7.6,c:\"#f59e0b\"},\n{l:\"Gemini API (Web:ON)\",v:7.3,c:\"#f59e0b\"},{l:\"ChatGPT Interface (Web:OFF)\",v:6.3,c:\"#6366f1\"},\n{l:\"Perplexity Interface (Web:ON)\",v:6.0,c:\"#00d4aa\"},{l:\"Perplexity Interface (Web:OFF)\",v:5.9,c:\"#00d4aa\"},\n{l:\"Gemini Interface (Web:OFF)\",v:5.7,c:\"#f59e0b\"},{l:\"ChatGPT API (Web:OFF)\",v:5.4,c:\"#6366f1\"},\n{l:\"ChatGPT API (Web:ON)\",v:5.0,c:\"#6366f1\"}];\nconst ac=document.getElementById('avgChart');\nad.forEach(d=>{ac.innerHTML+=`<div class=\"bar-row\"><div class=\"bar-label\">${d.l}<\/div><div class=\"bar-track\"><div class=\"bar-fill\" style=\"width:${d.v\/12*100}%;background:${d.c}\">${d.v}<\/div><\/div><div class=\"bar-meta\">firms\/response<\/div><\/div>`;});\n\n\/\/ === CHART 7: FAME vs QUALITY SCATTER  ===\n(function(){\nconst el=document.getElementById('fameChart');\nif(!el) return;\nconst data=[\n  {n:\"Richard Harris\",rev:3300,ai:585,r:4.7},\n  {n:\"Adam Kutner\",rev:2700,ai:233,r:4.7},\n  {n:\"Naqvi\",rev:1300,ai:782,r:4.9},\n  {n:\"PT Law\",rev:932,ai:3,r:5.0},\n  {n:\"De Castroverde\",rev:785,ai:103,r:4.7},\n  {n:\"Van Law\",rev:662,ai:52,r:5.0},\n  {n:\"Ladah\",rev:498,ai:268,r:4.8},\n  {n:\"Dimopoulos\",rev:3300,ai:86,r:4.9},\n];\nconst W=700,H=380,ml=60,mr=100,mt=30,mb=40;\nconst cW=W-ml-mr,cH=H-mt-mb;\nconst maxRev=3500,maxAI=800;\nconst sx=v=>ml+Math.log10(Math.max(v,1))\/Math.log10(maxRev)*cW;\nconst sy=v=>mt+cH-(v\/maxAI*cH);\n\nlet svg=`<svg viewBox=\"0 0 ${W} ${H}\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" style=\"width:100%;max-width:1000px\">`;\nsvg+=`<text x=\"${W\/2}\" y=\"16\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Review volume (how much people talk about you) predicts AI mentions. Rating (how good you are) doesn't.<\/text>`;\n\/\/ Grid\nfor(let ai=0;ai<=800;ai+=200){\n  const y=sy(ai);\n  svg+=`<line x1=\"${ml}\" y1=\"${y}\" x2=\"${ml+cW}\" y2=\"${y}\" stroke=\"#2a2a3a\" stroke-width=\"0.5\"\/>`;\n  svg+=`<text x=\"${ml-8}\" y=\"${y+4}\" text-anchor=\"end\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${ai}<\/text>`;\n}\nfor(let rv of [100,500,1000,3000]){\n  const x=sx(rv);\n  svg+=`<line x1=\"${x}\" y1=\"${mt}\" x2=\"${x}\" y2=\"${mt+cH}\" stroke=\"#2a2a3a\" stroke-width=\"0.5\"\/>`;\n  svg+=`<text x=\"${x}\" y=\"${mt+cH+14}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${rv>=1000?(rv\/1000)+'k':rv}<\/text>`;\n}\nsvg+=`<text x=\"${ml+cW\/2}\" y=\"${H-4}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Google review count (log scale)<\/text>`;\nsvg+=`<text x=\"14\" y=\"${mt+cH\/2}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\" transform=\"rotate(-90,14,${mt+cH\/2})\">Total AI mentions<\/text>`;\n\ndata.forEach(d=>{\n  const x=sx(d.rev),y=sy(d.ai);\n  const isHigh=d.ai>100;\n  const col=isHigh?'#22C55E':'#EF4444';\n  svg+=`<circle cx=\"${x}\" cy=\"${y}\" r=\"7\" fill=\"${col}\" opacity=\"0.8\"\/>`;\n  const anchor=x>ml+cW*0.6?\"end\":\"start\";\n  const dx=anchor===\"end\"?-12:12;\n  svg+=`<text x=\"${x+dx}\" y=\"${y+4}\" text-anchor=\"${anchor}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${d.n}<\/text>`;\n});\n\n\/\/ Callout arrows for key points\nsvg+=`<\/svg>`;\nel.innerHTML=svg;\n})();\n\/\/ === CHART 8: STABILITY BY CHANNEL  ===\n(function(){\nconst el=document.getElementById('stabilityChart');\nif(!el) return;\nconst data=[\n  {l:\"Gemini User Interface Web:OFF\",v:38.8,c:\"#22C55E\"},\n  {l:\"Perplexity API Web:OFF\",v:31.9,c:\"#00D4AA\"},\n  {l:\"Perplexity User Interface Web:ON\",v:28.7,c:\"#00D4AA\"},\n  {l:\"Perplexity User Interface Web:OFF\",v:28.4,c:\"#00D4AA\"},\n  {l:\"Gemini API Web:ON\",v:28.6,c:\"#F59E0B\"},\n  {l:\"ChatGPT User Interface Web:OFF\",v:24.7,c:\"#6366F1\"},\n  {l:\"Google AI Mode\",v:22.3,c:\"#FF6B35\"},\n  {l:\"Gemini User Interface Web:ON\",v:21.4,c:\"#F59E0B\"},\n  {l:\"ChatGPT User Interface Web:ON\",v:14.3,c:\"#6366F1\"},\n  {l:\"ChatGPT API Web:OFF\",v:8.8,c:\"#EF4444\"},\n  {l:\"ChatGPT API Web:ON\",v:6.5,c:\"#EF4444\"},\n];\nconst W=700,H=250,ml=170,mr=60,barH=11,gap=6;\nlet svg=`<svg viewBox=\"0 0 ${W} ${H}\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" style=\"width:100%;max-width:100%\">`;\nsvg+=`<text x=\"${W\/2}\" y=\"16\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Pairwise set overlap between repeat runs (higher = more stable)<\/text>`;\n\/\/ Mean line\nconst mean=23.8;\nconst bW=W-ml-mr;\nconst meanX=ml+mean\/50*bW;\nsvg+=`<line x1=\"${meanX}\" y1=\"25\" x2=\"${meanX}\" y2=\"${28+data.length*(barH+gap)}\" stroke=\"#8888a0\" stroke-width=\"1\" stroke-dasharray=\"4,4\"\/>`;\nsvg+=`<text x=\"${meanX}\" y=\"${28+data.length*(barH+gap)+14}\" text-anchor=\"middle\" fill=\"#8888a0\" font-size=\"11\" font-family=\"var(--theme-font-family)\">Mean = 23.8%<\/text>`;\n\ndata.forEach((d,i)=>{\n  const y=28+i*(barH+gap);\n  const w=d.v\/50*bW;\n  svg+=`<text x=\"${ml-8}\" y=\"${y+barH\/2+4}\" text-anchor=\"end\" fill=\"#e4e4ef\" font-size=\"10\" font-family=\"var(--theme-font-family)\">${d.l}<\/text>`;\n  svg+=`<rect x=\"${ml}\" y=\"${y}\" width=\"${w}\" height=\"${barH}\" fill=\"${d.c}\" rx=\"3\" opacity=\"0.85\"\/>`;\n  svg+=`<text x=\"${ml+w+6}\" y=\"${y+barH\/2+4}\" fill=\"#e4e4ef\" font-size=\"11\" font-family=\"var(--theme-font-family)\">${d.v}%<\/text>`;\n});\nsvg+=`<\/svg>`;\nel.innerHTML=svg;\n})();\n<\/script>\n\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>We asked 3 AI models the same 124 questions about personal injury lawyers across 3 channels \u2014 prompting AI interfaces directly, querying APIs, and Google Search AI. Here&#8217;s what they got right, and where they wildly disagree.<\/p>\n","protected":false},"author":1,"featured_media":264,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5],"tags":[],"class_list":["post-114","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-visibility"],"blocksy_meta":[],"_links":{"self":[{"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/posts\/114","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/comments?post=114"}],"version-history":[{"count":100,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/posts\/114\/revisions"}],"predecessor-version":[{"id":270,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/posts\/114\/revisions\/270"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/media\/264"}],"wp:attachment":[{"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/media?parent=114"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/categories?post=114"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/promptlocal.ai\/blog\/wp-json\/wp\/v2\/tags?post=114"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}