{"id":3902,"date":"2026-11-12T02:47:47","date_gmt":"2026-11-12T02:47:47","guid":{"rendered":"https:\/\/www.getoceana.com\/marketing\/?page_id=3902"},"modified":"2026-11-12T20:14:17","modified_gmt":"2026-11-12T20:14:17","slug":"why-ai-recommends-your-competitors-and-skips-you","status":"publish","type":"page","link":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/","title":{"rendered":"Why AI Recommends Your Competitors and Skips You"},"content":{"rendered":"\t\t<div data-elementor-type=\"wp-page\" data-elementor-id=\"3902\" class=\"elementor elementor-3902\">\n\t\t\t\t\t\t<section class=\"elementor-section elementor-top-section elementor-element elementor-element-3b1eda5 elementor-section-full_width elementor-section-height-default elementor-section-height-default\" data-id=\"3b1eda5\" data-element_type=\"section\" data-e-type=\"section\">\n\t\t\t\t\t\t<div class=\"elementor-container elementor-column-gap-default\">\n\t\t\t\t\t<div class=\"elementor-column elementor-col-100 elementor-top-column elementor-element elementor-element-5a5ff87\" data-id=\"5a5ff87\" data-element_type=\"column\" data-e-type=\"column\">\n\t\t\t<div class=\"elementor-widget-wrap elementor-element-populated\">\n\t\t\t\t\t\t<div class=\"elementor-element elementor-element-7c0cf32 elementor-widget elementor-widget-html\" data-id=\"7c0cf32\" data-element_type=\"widget\" data-e-type=\"widget\" data-widget_type=\"html.default\">\n\t\t\t\t<div class=\"elementor-widget-container\">\n\t\t\t\t\t<style>\r\n@import url('https:\/\/fonts.googleapis.com\/css2?family=Prompt:wght@300;500;600;700&family=Karla:wght@400;600;700&display=swap');\r\n.oce-article{--oce-blue:#3182CE;--oce-blue-d:#2B6CB0;--oce-ink:#1A202C;--oce-ink2:#2D3748;--oce-grey:#4A5568;--oce-grey2:#718096;--oce-tint:#EDF2F7;--oce-bg:#F7FAFC;font-family:Karla,-apple-system,BlinkMacSystemFont,'Segoe UI',sans-serif;color:var(--oce-ink);font-size:18px;line-height:1.6;-webkit-font-smoothing:antialiased;}\r\n.oce-article *,.oce-article *::before,.oce-article *::after{box-sizing:border-box;}\r\n\/* ===== HERO PHOTO: swap the URL below to change the hero image ===== *\/\r\n.oce-article{--oce-hero-img:url('https:\/\/www.getoceana.com\/wp-content\/uploads\/2022\/12\/Untitled-design-28.jpg');--oce-hero-overlay:.65;\r\n  --oce-author-img:url('https:\/\/www.getoceana.com\/wp-content\/uploads\/2022\/12\/Scott-Kath-1.jpg');}\r\n.oce-hero{position:relative;padding:180px 24px 70px;text-align:left;overflow:hidden;background-color:#2C5282;background-image:linear-gradient(135deg,#2C5282 0%,#3182CE 55%,#4299E1 100%);}\r\n.oce-hero-photo{position:absolute;inset:0;background-image:var(--oce-hero-img);background-size:cover;background-position:50% 50%;background-repeat:no-repeat;z-index:1;}\r\n.oce-hero-overlay{position:absolute;inset:0;background-color:#2B6CB0;opacity:var(--oce-hero-overlay);z-index:2;}\r\n.oce-hero-sheen{position:absolute;inset:0;z-index:3;background:radial-gradient(ellipse at 75% 115%,rgba(255,255,255,.22) 0%,rgba(255,255,255,0) 60%);pointer-events:none;}\r\n.oce-hero-inner{position:relative;z-index:4;max-width:1140px;margin:0 auto;}\r\n.oce-eyebrow{font-family:Karla,sans-serif;font-size:14px;line-height:1.6;font-weight:700;letter-spacing:1.5px;text-transform:uppercase;color:rgba(255,255,255,.82);margin:0 0 22px;}\r\n.oce-h1{font-family:Prompt,sans-serif;font-size:80px;line-height:80px;letter-spacing:-2px;font-weight:700;color:#fff;margin:0;}\r\n.oce-hero-sub{font-family:Karla,sans-serif;font-size:18px;line-height:28.8px;font-weight:700;color:#fff;margin:28px 0 0;max-width:820px;}\r\n.oce-deckwrap{background:#fff;padding:48px 24px 0;}\r\n.oce-deck{max-width:1140px;margin:0 auto;font-family:Prompt,sans-serif;font-size:22px;line-height:1.5;font-weight:500;color:var(--oce-ink);text-align:center;}\r\n.oce-body{background:#fff;padding:40px 24px 90px;}\r\n.oce-wrap{max-width:1140px;margin:0 auto;}\r\n.oce-h3{font-family:Prompt,sans-serif;font-size:30px;line-height:36px;font-weight:700;color:var(--oce-ink);margin:56px 0 20px;}\r\n.oce-h3:first-child{margin-top:0;}\r\n.oce-h4{font-family:Prompt,sans-serif;font-size:22px;line-height:30px;font-weight:500;color:var(--oce-ink);margin:38px 0 12px;}\r\n.oce-p{font-size:18px;line-height:28.8px;color:var(--oce-ink);margin:0 0 32px;}\r\n.oce-lead{font-weight:700;color:var(--oce-ink);}\r\n.oce-article a{color:var(--oce-blue);text-decoration:none;border-bottom:1px solid rgba(49,130,206,.35);transition:all .15s ease;}\r\n.oce-article a:hover{color:var(--oce-blue-d);border-bottom-color:var(--oce-blue-d);}\r\n.oce-ul,.oce-ol{margin:0 0 32px;padding-left:24px;}\r\n.oce-ul li,.oce-ol li{font-size:18px;line-height:28.8px;color:var(--oce-ink);margin-bottom:12px;padding-left:6px;}\r\n.oce-ul{list-style:none;padding-left:0;}\r\n.oce-ul li{position:relative;padding-left:28px;}\r\n.oce-ul li::before{content:\"\";position:absolute;left:0;top:11px;width:8px;height:8px;border-radius:50%;background:var(--oce-blue);}\r\n.oce-ol{counter-reset:oce;list-style:none;padding-left:0;}\r\n.oce-ol li{position:relative;padding-left:40px;counter-increment:oce;}\r\n.oce-ol li::before{content:counter(oce);position:absolute;left:0;top:2px;width:26px;height:26px;border-radius:50%;background:var(--oce-tint);color:var(--oce-blue-d);font-size:14px;font-weight:700;display:flex;align-items:center;justify-content:center;}\r\n.oce-tablewrap{overflow-x:auto;margin:0 0 36px;border-radius:6px;border:1px solid var(--oce-tint);}\r\n.oce-table{width:100%;border-collapse:collapse;font-size:17px;}\r\n.oce-table th{background:var(--oce-blue);color:#fff;font-family:Prompt,sans-serif;font-weight:700;font-size:14px;letter-spacing:.5px;text-transform:uppercase;text-align:left;padding:16px 20px;}\r\n.oce-table td{padding:16px 20px;border-top:1px solid var(--oce-tint);color:var(--oce-ink);}\r\n.oce-table tbody tr:nth-child(even){background:var(--oce-bg);}\r\n.oce-pre{background:var(--oce-ink);color:#E2E8F0;border-radius:6px;padding:28px 30px;overflow-x:auto;margin:0 0 36px;border-left:5px solid var(--oce-blue);}\r\n.oce-article code{background:var(--oce-tint);color:var(--oce-ink2);padding:2px 6px;border-radius:3px;font-size:16px;}\r\n.oce-article .oce-pre code{font-family:ui-monospace,SFMono-Regular,Menlo,Consolas,monospace;font-size:14.5px;line-height:1.75;white-space:pre;color:#E2E8F0;background:transparent;padding:0;border-radius:0;display:block;}\r\n.oce-sources li{font-size:16px;line-height:26px;color:var(--oce-grey);}\r\n.oce-sources li::before{background:var(--oce-grey2);top:9px;}\r\n.oce-cta{background:var(--oce-tint);padding:80px 24px;text-align:center;}\r\n.oce-cta-inner{max-width:760px;margin:0 auto;}\r\n.oce-cta h3{font-family:Prompt,sans-serif;font-size:50px;line-height:56px;font-weight:700;color:var(--oce-ink);margin:0 0 20px;}\r\n.oce-cta p{font-size:18px;line-height:28.8px;color:var(--oce-grey);margin:0 0 36px;}\r\n.oce-btns{display:flex;gap:16px;justify-content:center;flex-wrap:wrap;}\r\n.oce-btn{display:inline-block;font-family:Karla,sans-serif;font-size:16px;line-height:16px;font-weight:600;text-transform:uppercase;letter-spacing:.5px;padding:20px 50px;border-radius:6px;background:var(--oce-blue);color:#fff!important;border:0!important;transition:background .18s ease;}\r\n.oce-btn:hover{background:var(--oce-blue-d);color:#fff!important;}\r\n.oce-btn-alt{background:#fff;color:var(--oce-blue)!important;padding:20px 40px;border-radius:5px;}\r\n.oce-btn-alt:hover{background:#fff;color:var(--oce-blue-d)!important;}\r\n@media(max-width:1024px){.oce-h1{font-size:56px;line-height:58px;letter-spacing:-1.5px;}.oce-hero{padding:130px 24px 60px;}.oce-cta h3{font-size:38px;line-height:44px;}}\r\n@media(max-width:767px){.oce-h1{font-size:38px;line-height:42px;letter-spacing:-1px;}.oce-hero{padding:90px 20px 48px;}.oce-h3{font-size:26px;line-height:32px;margin:44px 0 16px;}.oce-h4{font-size:20px;line-height:28px;}.oce-deck{font-size:19px;}.oce-article,.oce-p,.oce-ul li,.oce-ol li{font-size:17px;}.oce-p,.oce-ul li,.oce-ol li{line-height:27px;}.oce-body{padding:32px 20px 64px;}.oce-cta{padding:56px 20px;}.oce-cta h3{font-size:30px;line-height:36px;}.oce-btn,.oce-btn-alt{width:100%;padding:18px 24px;}.oce-pre{padding:20px;}.oce-pre code{font-size:13px;}}\r\n\r\n.oce-chart{margin:0 0 40px;padding:30px 32px 22px;background:var(--oce-bg);border:1px solid var(--oce-tint);border-radius:8px;}\r\n.oce-ctitle{font-family:Prompt,sans-serif;font-size:20px;line-height:27px;font-weight:700;color:var(--oce-ink);margin:0 0 22px;}\r\n.oce-ctitle{position:relative;}\r\n.oce-ref{display:inline-flex;align-items:center;justify-content:center;min-width:21px;height:21px;padding:0 5px;margin-left:9px;border-radius:11px;background:#fff;border:1px solid #CBD5E0!important;color:var(--oce-grey)!important;font-family:Karla,sans-serif;font-size:12px;font-weight:700;line-height:1;vertical-align:4px;text-decoration:none;transition:all .15s ease;}\r\n.oce-ref:hover{background:var(--oce-blue);border-color:var(--oce-blue)!important;color:#fff!important;}\r\n.oce-cnote{font-size:14px;line-height:21px;color:var(--oce-grey);margin:18px 0 0;}\r\n.oce-bars{display:flex;flex-direction:column;gap:12px;}\r\n.oce-barrow{display:flex;align-items:center;gap:14px;}\r\n.oce-barlab{flex:0 0 33%;font-size:15px;line-height:20px;color:var(--oce-ink2);}\r\n.oce-bartrack{flex:1 1 auto;background:#E2E8F0;border-radius:4px;height:30px;overflow:hidden;display:block;}\r\n.oce-bar{display:block;height:100%;background:var(--oce-blue);border-radius:4px;}\r\n.oce-bar--accent{background:var(--oce-blue-d);}\r\n.oce-bar--mute{background:#A0AEC0;}\r\n.oce-barval{flex:0 0 62px;text-align:right;font-family:Prompt,sans-serif;font-size:17px;font-weight:700;color:var(--oce-ink);}\r\n.oce-barval--mute{color:var(--oce-grey2);}\r\n.oce-legend{display:flex;align-items:center;gap:20px;font-size:14px;color:var(--oce-grey);margin:-6px 0 18px;flex-wrap:wrap;}\r\n.oce-legitem{display:inline-flex;align-items:center;gap:7px;white-space:nowrap;}\r\n.oce-key{width:13px;height:13px;border-radius:3px;display:inline-block;flex:none;}\r\n.oce-key--a{background:var(--oce-blue);}\r\n.oce-key--b{background:#A0AEC0;}\r\n.oce-bars--grp{gap:18px;}\r\n.oce-grprow{display:flex;align-items:center;gap:14px;}\r\n.oce-grpbars{flex:1 1 auto;display:flex;flex-direction:column;gap:6px;}\r\n.oce-gline{display:flex;align-items:center;gap:14px;}\r\n.oce-gline .oce-bartrack{height:22px;}\r\n.oce-vs{display:flex;gap:56px;align-items:flex-start;justify-content:center;padding:6px 0 4px;}\r\n.oce-vsitem{flex:0 1 270px;text-align:center;display:flex;flex-direction:column;}\r\n.oce-vsnum{font-family:Prompt,sans-serif;font-size:40px;line-height:44px;font-weight:700;color:var(--oce-blue);margin-bottom:10px;min-height:44px;flex:none;}\r\n.oce-vsitem--mute .oce-vsnum{color:var(--oce-grey2);}\r\n.oce-vsbar{height:160px;display:flex;align-items:flex-end;flex:none;}\r\n.oce-vsbar>span{display:block;width:100%;background:var(--oce-blue);border-radius:5px 5px 0 0;}\r\n.oce-vsitem--mute .oce-vsbar>span{background:#A0AEC0;}\r\n.oce-vslab{font-size:14px;line-height:20px;color:var(--oce-grey);margin-top:12px;}\r\n.oce-stats{display:grid;grid-template-columns:repeat(4,1fr);gap:14px;margin:0 0 40px;}\r\n.oce-stat{background:var(--oce-blue);border-radius:8px;padding:24px 20px;color:#fff;display:flex;flex-direction:column;}\r\n.oce-statnum{font-family:Prompt,sans-serif;font-size:40px;line-height:42px;font-weight:700;letter-spacing:-1px;}\r\n.oce-statcap{font-size:15px;line-height:21px;margin-top:10px;color:rgba(255,255,255,.95);}\r\n.oce-statref{display:inline-block;align-self:flex-start;font-size:12px;line-height:17px;margin-top:auto;padding-top:12px;color:rgba(255,255,255,.78)!important;border-bottom:1px solid rgba(255,255,255,.4)!important;text-decoration:none;}\r\n.oce-statref:hover{color:#fff!important;border-bottom-color:#fff!important;}\r\n@media(max-width:860px){.oce-stats{grid-template-columns:repeat(2,1fr);}}\r\n@media(max-width:767px){\r\n .oce-chart{padding:22px 18px 18px;}.oce-ctitle{font-size:18px;line-height:24px;}\r\n .oce-barrow,.oce-grprow{flex-direction:column;align-items:stretch;gap:6px;}\r\n .oce-barlab{flex:none;font-size:14px;font-weight:700;}\r\n .oce-barval{flex:none;text-align:left;font-size:16px;}\r\n .oce-gline{flex-direction:row;align-items:center;}\r\n .oce-gline .oce-barval{flex:0 0 54px;text-align:right;}\r\n .oce-vs{gap:18px;}.oce-vsnum{font-size:30px;line-height:34px;}.oce-vsbar{height:110px;}.oce-vslab{font-size:13px;}\r\n .oce-stats{grid-template-columns:1fr;}.oce-statnum{font-size:34px;line-height:36px;}\r\n}\r\n\r\n.oce-author{display:flex;gap:28px;align-items:flex-start;background:var(--oce-tint);border-radius:8px;padding:34px 36px;margin:0 0 40px;}\r\n.oce-ava{flex:none;width:118px;height:118px;border-radius:50%;background-image:var(--oce-author-img);background-size:230%;background-position:44% 7%;background-repeat:no-repeat;box-shadow:0 0 0 4px #fff,0 0 0 6px var(--oce-blue);}\r\n.oce-aname{font-family:Prompt,sans-serif;font-size:24px;line-height:30px;font-weight:700;color:var(--oce-ink);margin:0;}\r\n.oce-arole{font-family:Karla,sans-serif;font-size:13px;font-weight:700;letter-spacing:1.2px;text-transform:uppercase;color:var(--oce-blue-d);margin:7px 0 14px;}\r\n.oce-author--slim{align-items:center;padding:28px 36px;}\r\n.oce-author--slim .oce-arole{margin-bottom:18px;}\r\n.oce-abio{font-size:17px;line-height:26px;color:var(--oce-ink2);margin:0 0 16px;}\r\n.oce-abio:last-child{margin-bottom:0;}\r\n.oce-alinks{display:flex;gap:10px;flex-wrap:wrap;}\r\n.oce-alink{display:inline-block;font-size:14px;font-weight:600;padding:9px 18px;border-radius:5px;background:#fff;color:var(--oce-blue)!important;border:1px solid #fff!important;text-decoration:none;}\r\n.oce-alink:hover{background:var(--oce-blue);color:#fff!important;border-color:var(--oce-blue)!important;}\r\n.oce-srclist{counter-reset:src;list-style:none;padding-left:0;margin:0 0 32px;}\r\n.oce-srclist li{position:relative;padding-left:40px;counter-increment:src;font-size:16px;line-height:25px;color:var(--oce-grey);margin-bottom:11px;scroll-margin-top:90px;}\r\n.oce-srclist li::before{content:counter(src);position:absolute;left:0;top:2px;width:25px;height:25px;border-radius:50%;background:#fff;border:1px solid #CBD5E0;color:var(--oce-grey);font-size:12px;font-weight:700;display:flex;align-items:center;justify-content:center;}\r\n.oce-srclist li:target::before{background:var(--oce-blue);border-color:var(--oce-blue);color:#fff;}\r\n.oce-srclist li:target{color:var(--oce-ink);}\r\n.oce-refnote{font-size:15px;line-height:23px;color:var(--oce-grey);background:var(--oce-bg);border-left:4px solid var(--oce-blue);padding:14px 18px;border-radius:0 5px 5px 0;margin:0 0 36px;}\r\n@media(max-width:767px){\r\n .oce-author{flex-direction:column;gap:18px;padding:24px 20px;align-items:center;text-align:center;}\r\n .oce-ava{width:96px;height:96px;}\r\n .oce-alinks{justify-content:center;}\r\n .oce-aname{font-size:21px;line-height:27px;}.oce-abio{font-size:16px;line-height:25px;}\r\n}\r\n\r\n.oce-amore{display:flex;align-items:center;flex-wrap:wrap;gap:10px;margin:20px 0 0;padding-top:18px;border-top:1px solid #DCE4EC;font-size:14px;line-height:20px;}\r\n.oce-amorelab{margin-right:6px;font-family:Karla,sans-serif;font-size:12px;font-weight:700;letter-spacing:1.1px;text-transform:uppercase;color:var(--oce-grey2);}\r\n.oce-aquote{margin:2px 0 18px;padding:15px 22px;background:#fff;border-left:4px solid var(--oce-blue);border-radius:0 6px 6px 0;font-family:Prompt,sans-serif;font-size:18px;line-height:27px;font-weight:500;font-style:normal;color:var(--oce-ink);}\r\n@media(max-width:767px){.oce-aquote{font-size:17px;line-height:26px;padding:13px 16px;text-align:left;}}\r\n.oce-ico{display:inline-flex;align-items:center;justify-content:center;width:38px;height:38px;border-radius:50%;background:#fff;border:1px solid #DCE4EC!important;text-decoration:none!important;transition:background .15s ease,border-color .15s ease;}\r\n.oce-ico svg{width:19px;height:19px;fill:none;stroke:var(--oce-blue-d);stroke-width:1.7;stroke-linecap:round;stroke-linejoin:round;transition:stroke .15s ease;}\r\n.oce-ico:hover{background:var(--oce-blue);border-color:var(--oce-blue)!important;}\r\n.oce-ico:hover svg{stroke:#fff;}\r\n.oce-author .oce-arole{line-height:19px;}\r\n@media(max-width:767px){.oce-amore{justify-content:center;gap:8px 14px;}}\r\n<\/style>\r\n\r\n<div class=\"oce-article\">\r\n\r\n  <section class=\"oce-hero\">\r\n    <div class=\"oce-hero-photo\"><\/div>\r\n    <div class=\"oce-hero-overlay\"><\/div>\r\n    <div class=\"oce-hero-sheen\"><\/div>\r\n    <div class=\"oce-hero-inner\">\r\n      <h1 class=\"oce-h1\">Why AI Recommends Your Competitors and Skips You<\/h1>\r\n      <p class=\"oce-hero-sub\">The 2026 research on this now covers 704 homepages, 158,099 AI answers and several million local citations across a dozen separate studies, and it points one way.<\/p>\r\n    <\/div>\r\n  <\/section>\r\n\r\n  <section class=\"oce-deckwrap\">\r\n    <p class=\"oce-deck\">The assistant is counting what it can verify on your page, and most homepages give it very little to count.<\/p>\r\n  <\/section>\r\n\r\n  <section class=\"oce-body\">\r\n    <div class=\"oce-wrap\">\r\n<h3 class=\"oce-h3\">The short answer, before anything else<\/h3>\r\n<p class=\"oce-refnote\">Every chart on this page carries a numbered reference. The number links to the full study, sample size and publication date in the sources list at the foot of the page.<\/p>\r\n<div class=\"oce-stats\"><div class=\"oce-stat\"><div class=\"oce-statnum\">30%<\/div><div class=\"oce-statcap\">chance of being recommended at a homepage carrying around 60 checkable facts, against 5% at 15<\/div><a class=\"oce-statref\" href=\"#oce-src-1\">Orbit Media, November 2026<\/a><\/div><div class=\"oce-stat\"><div class=\"oce-statnum\">7%<\/div><div class=\"oce-statcap\">of B2B buyers notice a vendor in an AI answer because they recognize the name<\/div><a class=\"oce-statref\" href=\"#oce-src-2\">Semrush, July 2026<\/a><\/div><div class=\"oce-stat\"><div class=\"oce-statnum\">92%<\/div><div class=\"oce-statcap\">say AI has already shaped their vendor shortlist, 45% of them significantly<\/div><a class=\"oce-statref\" href=\"#oce-src-2\">Semrush, July 2026<\/a><\/div><div class=\"oce-stat\"><div class=\"oce-statnum\">74.9%<\/div><div class=\"oce-statcap\">of 10,000 US restaurants never appeared in Google's AI recommendations at all<\/div><a class=\"oce-statref\" href=\"#oce-src-8\">Local Falcon, June 2026<\/a><\/div><\/div>\r\n<p class=\"oce-p\">When a buyer asks ChatGPT, Gemini, Perplexity or Google's AI Mode to recommend a company like yours, the assistant works from what it can check. Specific, verifiable facts move the needle: named clients, real outcomes, credentials written as text, particular capabilities and equipment, the industries and towns you actually serve. Length does not help. Polish does not help. Self-description appears to work against you.<\/p>\r\n<p class=\"oce-p\">The clearest numbers on this landed on November 4, 2026, in research published by Andy Crestodina and the team at Orbit Media. Homepages carrying around 15 checkable facts were recommended roughly 5% of the time. At around 60 facts, that rose to roughly 30%. Word count showed no connection at all. Extra words helped only when those words carried facts.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Recommendation rate by number of checkable facts on the homepage<a class=\"oce-ref\" href=\"#oce-src-1\" title=\"Jump to source 1\">1<\/a><\/figcaption><div class=\"oce-vs\"><div class=\"oce-vsitem\"><div class=\"oce-vsnum\">30%<\/div><div class=\"oce-vsbar\"><span style=\"height:100.0%\"><\/span><\/div><div class=\"oce-vslab\">Homepage with around 60 checkable facts<\/div><\/div><div class=\"oce-vsitem oce-vsitem--mute\"><div class=\"oce-vsnum\">5%<\/div><div class=\"oce-vsbar\"><span style=\"height:16.7%\"><\/span><\/div><div class=\"oce-vslab\">Homepage with around 15 checkable facts<\/div><\/div><\/div><p class=\"oce-cnote\">Main analysis of 575 companies drawn from 704 collected homepages. The average homepage in the sample stated 31 facts.<\/p><\/figure>\r\n<p class=\"oce-p\">The typical homepage in their sample stated 31 facts. So the work between an average page and a recommended page is inventory work.<\/p>\r\n<h3 class=\"oce-h3\">What the homepage study actually did<\/h3>\r\n<p class=\"oce-p\">This matters because the methodology sets the limits of what you can claim from it.<\/p>\r\n<p class=\"oce-p\">The researchers collected 704 company homepages found through Google across four industries: web design agencies, IT managed service providers, CPA firms, and manufacturers. The main analysis removed the top quarter by Domain Authority, leaving 575 companies. That excluded top quarter was recommended 44% of the time, which is the honest caveat buried in the data: established authority wins recommendations on its own.<\/p>\r\n<p class=\"oce-p\">They wrote 48 buyer needs, phrased each one twice (once as a short keyword, once as a detailed request) for 96 prompts total. Each prompt ran five times through ChatGPT, Gemini, Google AI Mode and Perplexity in October 2026, in US English, collected through DataForSEO. That produced 1,916 answers and 8,771 recommendations, averaging 4.6 recommended companies per answer.<\/p>\r\n<p class=\"oce-p\">The outcome they measured was the chance of being recommended at least once across 40 answers, adjusted for word count, Domain Authority, industry and total fact count. The findings are correlational rather than causal, which the authors state plainly.<\/p>\r\n<p class=\"oce-p\">That authority figure reframes the whole exercise. If you are a small or mid-sized operator without the domain authority of a national brand, fact density is the lever available to you. It is the one input that does not require a decade of link acquisition.<\/p>\r\n<h3 class=\"oce-h3\">The categories that correlated, with the thresholds<\/h3>\r\n<p class=\"oce-p\">The study did not treat all facts as equal. Specific categories carried specific weight, and several had thresholds where the effect appeared.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Named clients and results.<\/span> The strongest single correlation in the study. Homepages naming five or more clients or client outcomes were recommended 30% of the time. Homepages naming none came in at 11%.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Credentials.<\/span> Three or more credentials correlated with a 25% recommendation rate, against 12% for pages showing none. A single certification showed no measurable effect, which is worth knowing before you put one badge in your footer and call it done.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Third-party recognition.<\/span> Two or more awards, rankings or ratings correlated with 22%, against 13% for pages with none.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Capabilities, equipment, facilities and tools.<\/span> Pages naming them: 25%. Pages that did not: 13%. For a service business, the raw material for this category already sits in your equipment and software list.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Industries served.<\/span> Naming six to nine industries outperformed naming fewer. Listing more than that did not help. There is a real ceiling here, which argues against the dump-everything approach.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Buyer alignment.<\/span> This finding deserves more attention than it has received. The study checked whether a homepage stated the specific things the buyer asked for. Around a quarter of all homepages in the sample stated zero facts that aligned with what buyers were asking for.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Recommendation rate by category of fact<a class=\"oce-ref\" href=\"#oce-src-1\" title=\"Jump to source 1\">1<\/a><\/figcaption><div class=\"oce-legend\"><span class=\"oce-legitem\"><span class=\"oce-key oce-key--a\"><\/span>Homepages that had it<\/span><span class=\"oce-legitem\"><span class=\"oce-key oce-key--b\"><\/span>Homepages that did not<\/span><\/div><div class=\"oce-bars oce-bars--grp\"><div class=\"oce-grprow\"><span class=\"oce-barlab\">Five or more named clients and results<\/span><span class=\"oce-grpbars\"><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">30%<\/span><\/span><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:32.74%\"><\/span><\/span><span class=\"oce-barval oce-barval--mute\">11%<\/span><\/span><\/span><\/div><div class=\"oce-grprow\"><span class=\"oce-barlab\">Three or more credentials<\/span><span class=\"oce-grpbars\"><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:74.40%\"><\/span><\/span><span class=\"oce-barval\">25%<\/span><\/span><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:35.71%\"><\/span><\/span><span class=\"oce-barval oce-barval--mute\">12%<\/span><\/span><\/span><\/div><div class=\"oce-grprow\"><span class=\"oce-barlab\">Capabilities, equipment and tools named<\/span><span class=\"oce-grpbars\"><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:74.40%\"><\/span><\/span><span class=\"oce-barval\">25%<\/span><\/span><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:38.69%\"><\/span><\/span><span class=\"oce-barval oce-barval--mute\">13%<\/span><\/span><\/span><\/div><div class=\"oce-grprow\"><span class=\"oce-barlab\">Two or more third-party awards or ratings<\/span><span class=\"oce-grpbars\"><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:65.48%\"><\/span><\/span><span class=\"oce-barval\">22%<\/span><\/span><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:38.69%\"><\/span><\/span><span class=\"oce-barval oce-barval--mute\">13%<\/span><\/span><\/span><\/div><\/div><p class=\"oce-cnote\">A single certification on its own showed no measurable effect.<\/p><\/figure>\r\n<p class=\"oce-p\">A quarter of companies are invisible for a reason that has nothing to do with technology. Their page does not mention the thing the buyer typed.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Stating what the buyer actually asked for<a class=\"oce-ref\" href=\"#oce-src-1\" title=\"Jump to source 1\">1<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">None of the buyer's requirements stated<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:21.01%\"><\/span><\/span><span class=\"oce-barval\">4%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Some requirements stated<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:57.77%\"><\/span><\/span><span class=\"oce-barval\">11%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Half or more stated<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--accent\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">17%<\/span><\/div><\/div><p class=\"oce-cnote\">Around a quarter of all homepages in the sample stated nothing at all that aligned with what buyers were asking for.<\/p><\/figure>\r\n<h3 class=\"oce-h3\">Self-praise correlated with fewer recommendations<\/h3>\r\n<p class=\"oce-p\">Over half the homepages studied called themselves best, leading or world-class. About three in four used words like innovative, seamless or passionate.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Recommendation rate by use of self-praise words<a class=\"oce-ref\" href=\"#oce-src-1\" title=\"Jump to source 1\">1<\/a><\/figcaption><div class=\"oce-vs\"><div class=\"oce-vsitem\"><div class=\"oce-vsnum\">16%<\/div><div class=\"oce-vsbar\"><span style=\"height:100.0%\"><\/span><\/div><div class=\"oce-vslab\">Homepages without the praise words<\/div><\/div><div class=\"oce-vsitem oce-vsitem--mute\"><div class=\"oce-vsnum\">10%<\/div><div class=\"oce-vsbar\"><span style=\"height:62.5%\"><\/span><\/div><div class=\"oce-vslab\">Homepages heaviest on best, leading, world-class<\/div><\/div><\/div><p class=\"oce-cnote\">Over half of all homepages studied called themselves best, leading or world-class.<\/p><\/figure>\r\n<p class=\"oce-p\">That gap is the uncomfortable part of this research, and it is correlational, so the mechanism stays unproven. The plausible reading is that praise words occupy the space facts would otherwise fill. A sentence spent asserting that you are trusted is a sentence that did not name a client, a certification, a piece of equipment or a town.<\/p>\r\n<p class=\"oce-p\">The replacement rule is simple to apply. Where a sentence describes how good you are, rewrite it as what you did, for whom, and when.<\/p>\r\n<h3 class=\"oce-h3\">The badge problem<\/h3>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">How a credential is displayed changes what the assistant can read<a class=\"oce-ref\" href=\"#oce-src-1\" title=\"Jump to source 1\">1<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">Badge or logo image only<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:46.99%\"><\/span><\/span><span class=\"oce-barval\">10%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Credential written as text only<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:70.49%\"><\/span><\/span><span class=\"oce-barval\">15%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Badge and text together<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--accent\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">19%<\/span><\/div><\/div><\/figure>\r\n<p class=\"oce-p\">Assistants read markup. A credential that exists on your site only inside an image file does not register with them. If you do one thing from this article, write out every certification, license, award and association membership that currently exists on your site only as a picture.<\/p>\r\n<h3 class=\"oce-h3\">Where the standard advice does not hold up<\/h3>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">FAQ sections.<\/span> In the sample, 22% of homepages carried an FAQ section, typically with six questions. FAQ blocks containing few facts correlated with a 12% recommendation rate. FAQ blocks containing many facts: 28%. The format itself added nothing beyond the facts inside it. An FAQ section is a container, and an empty container earns nothing.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Pricing.<\/span> Only 8% of homepages in the sample published prices, and the study detected no correlation between showing prices and being recommended. Price-related terms such as flat monthly price or no long-term contract did appear to help in quote-based industries. So the case for publishing numbers rests on buyer behavior, covered below, and the recommendation-side evidence stays neutral. This article will not overstate it.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Factors with no measurable effect in the study:<\/span> word count, testimonials, branded process names, statements about what the company does not do, company profile trivia (years in business, founding date, team size, ownership), and generic headings or navigation labels.<\/p>\r\n<p class=\"oce-p\">If your homepage leads with a branded process name and a founding date, the study suggests you are spending your most-read real estate on content that earned nothing.<\/p>\r\n<h3 class=\"oce-h3\">Your homepage is carrying almost all of the weight<\/h3>\r\n<p class=\"oce-p\">This is the structural finding most owners miss.<\/p>\r\n<p class=\"oce-p\">Citing earlier Cloudflare data across 74 websites, the study reports that AI bots request homepages roughly 15 times more often than any other page type. In their own analysis, two-thirds of what the AI said about a recommended company traced back to the homepage. Only 5% traced to an interior page, and they checked four per site (about, services, industries, work).<\/p>\r\n<p class=\"oce-p\">Homepage fact count alone moved recommendation likelihood from roughly 6% to 26%.<\/p>\r\n<p class=\"oce-p\">The practical consequence: your deep service pages are doing far less recommendation work than your homepage. Facts that live three clicks down are facts the assistant may never reach.<\/p>\r\n<h3 class=\"oce-h3\">Why it works this way: the retrieval layer<\/h3>\r\n<p class=\"oce-p\">The mechanism behind all of this was documented by Mahe Chen and Nick Koudas in a paper published to arXiv on October 10, 2026, covering 158,099 AI answers containing 468,037 citations.<\/p>\r\n<p class=\"oce-p\">Their central result comes from replaying the same questions through the engines' own APIs. Which pages the engine's search step retrieved explained 32% of the variation in whether a page got cited. Page properties explained 2.8%.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">What explains whether a page gets cited<a class=\"oce-ref\" href=\"#oce-src-3\" title=\"Jump to source 3\">3<\/a><\/figcaption><div class=\"oce-vs\"><div class=\"oce-vsitem\"><div class=\"oce-vsnum\">32%<\/div><div class=\"oce-vsbar\"><span style=\"height:100.0%\"><\/span><\/div><div class=\"oce-vslab\">Which pages the engine's own search retrieved<\/div><\/div><div class=\"oce-vsitem oce-vsitem--mute\"><div class=\"oce-vsnum\">2.8%<\/div><div class=\"oce-vsbar\"><span style=\"height:8.8%\"><\/span><\/div><div class=\"oce-vslab\">Properties of the page itself<\/div><\/div><\/div><p class=\"oce-cnote\">158,099 AI answers containing 468,037 citations. Share of the variation in citation explained by each factor.<\/p><\/figure>\r\n<p class=\"oce-p\">Matching the buyer's actual question is roughly ten times more consequential than the markup most SEO advice prioritizes.<\/p>\r\n<p class=\"oce-p\">The paper is specific about what that means in practice. Comparing pages within the same website, the pages whose titles and paragraphs matched the questions people genuinely ask were cited more. FAQ blocks, structured data, summary boxes and page speed appeared effective only when comparing one website against another, which is a far weaker claim and easily confounded by site quality. Every page ChatGPT cited had been retrieved by its own search step first. And the sub-queries ChatGPT writes for itself often differ from what Google and Bing return, so your Google rankings are an imperfect proxy for what the assistant sees.<\/p>\r\n<p class=\"oce-p\">Two further findings from the same paper are worth holding onto. Model knowledge plays a real part: ChatGPT includes brands it already knows in its own searches, prefers websites it knows, and sometimes names brands with no supporting page at all. In controlled experiments, brand names alone reordered product rankings when the underlying product facts were identical. And the authors report that their pattern replicated in a market of small B2B vendors, so the finding extends well beyond enterprise software.<\/p>\r\n<p class=\"oce-p\">Brand familiarity is real and it compounds. Fact density is the part you can change this quarter.<\/p>\r\n<h3 class=\"oce-h3\">The finding almost nobody is talking about: the results move<\/h3>\r\n<p class=\"oce-p\">Every study here independently found that AI recommendations are unstable, and this changes how you should measure anything.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">The homepage study:<\/span> only 4% to 18% of recommended companies appeared in all five runs of the same prompt.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">BrightLocal<\/span> , in a September 2026 study covering around 200,000 localized non-branded prompts and 1.97 million citations across roughly 1,300 locations: the average overlap between repeat result sets was 20% to 33% (ChatGPT 23%, AI Mode 21%, AI Overviews 33%). Between 71% and 80% of businesses appeared in half of attempts or fewer. A business seen once reappeared later roughly half the time.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Steady Demand<\/span> , led by Google Business Profile Diamond Product Expert Ben Fisher, published its AI Citation Ledger on August 6, 2026: 14,472 citations classified from 1,487 local-service queries across 50 US metros and 10 verticals. Repeating an identical query, Gemini returned the same top business 7.9% of the time. Google's local pack returned the same top listing 90.2% of the time. Cited-domain overlap on an immediate repeat was 46% at best, falling to 41% after about three and a half hours and 38% the next day.<\/p>\r\n<p class=\"oce-p\">The same team found Gemini and ChatGPT shared only 8.3% of cited domains on identical queries, and named the same top business on 4.2% of them.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Ask the same question twice and the answer changes<a class=\"oce-ref\" href=\"#oce-src-5\" title=\"Jump to source 5\">5<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">Google's local pack returns the same top listing<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">90.2%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Gemini returns the same top business<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:7.82%\"><\/span><\/span><span class=\"oce-barval\">7.9%<\/span><\/div><\/div><p class=\"oce-cnote\">14,472 citations from 1,487 local-service queries across 50 US metros.<\/p><\/figure>\r\n<p class=\"oce-p\">Three operational consequences follow:<\/p>\r\n<ol class=\"oce-ol\"><li>A single check tells you nothing. If you ask ChatGPT once whether it recommends you and it does, that result had somewhere between a 4% and a 23% chance of repeating.<\/li><li>Measure as a rate, across runs and engines. Five prompts, four assistants, three runs each, logged monthly. The number that matters is how often you appear across the full set of checks.<\/li><li>The ground itself shifts. In that same dataset, ChatGPT's Reddit citation share for local queries went from 41.7% to 0.0% overnight on August 20, 2026. A tactic built on one citation source can disappear in a day. Facts on your own property do not.<\/li><\/ol>\r\n<h3 class=\"oce-h3\">If you are a local service business, the picture is different<\/h3>\r\n<p class=\"oce-p\">Most of the research above studied B2B services and manufacturers. For local and home-service companies, the source mix changes significantly, and it is worth being precise about what applies to whom.<\/p>\r\n<p class=\"oce-p\">Across ChatGPT, Google AI Mode and Google AI Overviews, the September 2026 BrightLocal data put Google Business Profile at about 28.5% of all local citations, the single largest source.<\/p>\r\n<p class=\"oce-p\">So for local recommendations, two assets are doing the work: a complete Google Business Profile, and your own website. Everything else is a rounding error.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Where AI answers about local businesses get their information<a class=\"oce-ref\" href=\"#oce-src-4\" title=\"Jump to source 4\">4<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">The business's own website<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">42%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Google Business Profile<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:60.59%\"><\/span><\/span><span class=\"oce-barval\">28.5%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Yelp<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:20.20%\"><\/span><\/span><span class=\"oce-barval\">9.5%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Facebook<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:4.68%\"><\/span><\/span><span class=\"oce-barval\">2.2%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Tripadvisor<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:3.83%\"><\/span><\/span><span class=\"oce-barval\">1.8%<\/span><\/div><\/div><p class=\"oce-cnote\">Around 200,000 localized prompts and 1.97 million citations across roughly 1,300 locations. Business websites supplied 108,000 of the 116,000 unique source domains cited, or 93%.<\/p><\/figure>\r\n<p class=\"oce-p\">A second study from the same team, published August 22, 2026 and covering 7,990 Google queries and 86,645 citation references, adds a split that changes the work order. On local-intent queries, AI Overviews cited business websites 70.7% of the time, while AI Mode cited them 12.8% of the time and cited Google Maps 79.8% of the time. On informational queries, AI Mode cited business sites 42.3% of the time.<\/p>\r\n<p class=\"oce-p\">Read together: AI Overviews reward your website, AI Mode reads your Google Business Profile, and both need attention because they reward close to opposite things.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Google's two AI surfaces reward opposite things<a class=\"oce-ref\" href=\"#oce-src-5\" title=\"Jump to source 5\">5<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">AI Overviews citing business websites<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:79.10%\"><\/span><\/span><span class=\"oce-barval\">70.7%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">AI Mode citing Google Maps<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">79.8%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">AI Mode citing business websites<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:14.32%\"><\/span><\/span><span class=\"oce-barval\">12.8%<\/span><\/div><\/div><p class=\"oce-cnote\">7,990 queries and 86,645 citation references, local-intent queries.<\/p><\/figure>\r\n<p class=\"oce-p\">Three more figures from BrightLocal that should shape expectations:<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">The map still outperforms the assistants.<\/span> Tracked businesses were mentioned by Google Maps 66% of the time, against 38% for AI Overviews, 33% for ChatGPT and 32% for AI Mode. Maps was consistent across attempts 99% of the time.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Geography behaves strangely.<\/span> The radius covering 90% of recommended results was about 24 miles (38km) for AI Overviews, around 40 miles (64km) for AI Mode, and 178 miles (287km) for ChatGPT. ChatGPT reaches across a whole region in a way the map never would, which cuts both ways: your competitive set is wider than your service area, and so is your opportunity.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Service-area businesses start behind.<\/span> Storefronts were mentioned 33.6% of the time against 29.4% for service-area businesses. Mobile and dispatch-based operators start with a structural disadvantage, and they have the most to gain from the fact work.<\/p>\r\n<p class=\"oce-p\">Separately, in early 2026, 45% of consumers told BrightLocal they had used ChatGPT or another generative AI tool for local business recommendations.<\/p>\r\n<h3 class=\"oce-h3\">Being the biggest name in town offers no protection<\/h3>\r\n<p class=\"oce-p\">The most reassuring finding of 2026 came from Local Falcon's Restaurant AI Visibility Index, published June 12, 2026. The team ran 10,000 US restaurants across all 50 states and Washington, D.C. through both Google Maps and Google AI Overviews using May 2026 data, checking each business from nine grid points around its own location using its primary cuisine category, all at midday.<\/p>\r\n<p class=\"oce-p\">The headline result reverses the assumption most owners carry into this:<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Independents outperform chains in AI recommendations<a class=\"oce-ref\" href=\"#oce-src-8\" title=\"Jump to source 8\">8<\/a><\/figcaption><div class=\"oce-legend\"><span class=\"oce-legitem\"><span class=\"oce-key oce-key--a\"><\/span>Appears on Google Maps<\/span><span class=\"oce-legitem\"><span class=\"oce-key oce-key--b\"><\/span>Appears in AI<\/span><\/div><div class=\"oce-bars oce-bars--grp\"><div class=\"oce-grprow\"><span class=\"oce-barlab\">Independent restaurants<\/span><span class=\"oce-grpbars\"><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:83.67%\"><\/span><\/span><span class=\"oce-barval\">79.0%<\/span><\/span><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:30.82%\"><\/span><\/span><span class=\"oce-barval oce-barval--mute\">29.1%<\/span><\/span><\/span><\/div><div class=\"oce-grprow\"><span class=\"oce-barlab\">Chains with five or more locations<\/span><span class=\"oce-grpbars\"><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">84.3%<\/span><\/span><span class=\"oce-gline\"><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:14.72%\"><\/span><\/span><span class=\"oce-barval oce-barval--mute\">13.9%<\/span><\/span><\/span><\/div><\/div><p class=\"oce-cnote\">10,000 US restaurants across all 50 states, nine grid points each, May 2026 data.<\/p><\/figure>\r\n\r\n<p class=\"oce-p\">Independents appeared in Google's AI recommendations more than twice as often as chains, even though the chains held the advantage on Maps. Local Falcon also reports that the six chains with more than 100 locations in the study went missing from Google's AI recommendations as often as ordinary restaurants, and sometimes more often.<\/p>\r\n<p class=\"oce-p\">The rest of that study sets expectations properly:<\/p>\r\n<ul class=\"oce-ul\"><li>74.9% of the 10,000 restaurants never appeared in Google's AI recommendations at a single one of their nine grid points, against 19.6% that never appeared on Google Maps.<\/li><li>A restaurant with more than 1,000 reviews was still left out 70.9% of the time.<\/li><li>5.4% of the restaurants ChatGPT recommended were rated below 3.5 stars, even though the prompt asked for highly rated places.<\/li><li>Yelp took 14.1% of citations in that study, more than TripAdvisor, Facebook, Reddit and Instagram combined.<\/li><li>In a compilation of Local Falcon's figures, the top 10% of restaurants captured 74.5% of all AI visibility, against 54% on Google Maps.<\/li><\/ul>\r\n<p class=\"oce-p\">Two conclusions for anyone who is not a national brand. The concentration is severe, so a large majority of the field is currently invisible and the opening is real. And brand size is not the thing the assistants are rewarding, which runs opposite to how the local pack has behaved for years.<\/p>\r\n<p class=\"oce-p\">Scale also sets the baseline. SOCi's 2026 Local Visibility Index, covering more than 350,000 US locations across 2,751 multi-location brands, found ChatGPT recommended 1.2% of locations, Gemini 11% and Perplexity 7.4%, against 35.9% for the same businesses in Google's local 3-pack. Those figures describe multi-location brands rather than independents, so read them as the scale of the gap rather than as your own odds.<\/p>\r\n<h3 class=\"oce-h3\">Reviews, and the number nobody agrees on<\/h3>\r\n<p class=\"oce-p\">Zarla's AI Visibility Index, collected on August 24, 2026, asked ChatGPT with search enabled and Google AI Mode about 1,851 top-ranked US local businesses across ten trades and ten cities.<\/p>\r\n<ul class=\"oce-ul\"><li>ChatGPT named 35.6% of the businesses (659 of 1,851). Google AI Mode named 18.4% (330 of 1,794).<\/li><li>Only 181 businesses were named by both engines. 478 were named by ChatGPT alone and 149 by AI Mode alone.<\/li><li>Every business that got named already ranked in Google's top 20 for its trade and city, as reported from that index. Standard local SEO is the entry requirement, and the AI layer sits on top of it.<\/li><\/ul>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Median Google reviews, businesses named against businesses skipped<a class=\"oce-ref\" href=\"#oce-src-9\" title=\"Jump to source 9\">9<\/a><\/figcaption><div class=\"oce-vs\"><div class=\"oce-vsitem\"><div class=\"oce-vsnum\">444<\/div><div class=\"oce-vsbar\"><span style=\"height:100.0%\"><\/span><\/div><div class=\"oce-vslab\">Median reviews, businesses the assistants named<\/div><\/div><div class=\"oce-vsitem oce-vsitem--mute\"><div class=\"oce-vsnum\">192<\/div><div class=\"oce-vsbar\"><span style=\"height:43.2%\"><\/span><\/div><div class=\"oce-vslab\">Median reviews, businesses they skipped<\/div><\/div><\/div><p class=\"oce-cnote\">1,851 top-ranked US local businesses across ten trades and ten cities, collected August 24, 2026. Star ratings across the two groups were 4.9 and 4.8.<\/p><\/figure>\r\n<p class=\"oce-p\">Now the part that matters more than any single figure. The review-volume finding replicates everywhere. The number attached to it does not.<\/p>\r\n<ul class=\"oce-ul\"><li>Zarla, 1,851 US local businesses, August 2026: a median of 444 reviews against 192.<\/li><li>Insites, 10,000 US local businesses run through ChatGPT and Perplexity: recommended businesses averaged 133.4 Google reviews, ignored ones averaged 10.7, a twelvefold gap.<\/li><li>SOCi's 2026 Local Visibility Index: locations ChatGPT recommended averaged 4.3 stars, which sat only 0.1 above the 4.2 average across every location in the study.<\/li><\/ul>\r\n<p class=\"oce-p\">Three samples, three very different review counts, one consistent direction. Review volume separates the named from the skipped, and star rating does very little of that work. Use it as a direction of travel for your review program, and treat any specific review threshold quoted to you as a cutoff with suspicion, because this body of work does not support one.<\/p>\r\n<p class=\"oce-p\">A related pattern comes from Cheers, drawn from 34,017 rated business appearances across 8,523 Google AI Mode and ChatGPT answers collected between August 31 and September 20, 2026: in answers naming three or more rated businesses, the business named first tended to carry the most reviews. Cheers states plainly that this describes associations in their sample rather than how any engine ranks businesses.<\/p>\r\n<p class=\"oce-p\">One more detail from the Insites analysis has direct consequences for content planning. When ChatGPT cites a source, they found it four times more likely to be Reddit or Wikipedia than the press.<\/p>\r\n<p class=\"oce-p\">The practical read: a steady flow of recent reviews does more for your AI visibility than the last tenth of a star, and the order in which you are named may track review count too.<\/p>\r\n<h3 class=\"oce-h3\">Ads have already arrived in local AI answers<\/h3>\r\n<p class=\"oce-p\">This is the development with the shortest fuse, and very few people in local marketing are discussing it yet.<\/p>\r\n<p class=\"oce-p\">In late August 2026, Steady Demand ran 1,200 local-intent prompts across 10 verticals, six US markets and four prompt templates, alongside a 200-prompt general-commercial control group carrying no location. The study was published August 21, 2026, with 95% Wilson confidence intervals on the rates.<\/p>\r\n<p class=\"oce-p\">Ads rendered on 58.25% of local-intent runs (confidence interval 55.44% to 61.01%), against 40.50% for the control with no location attached. Local intent drew ads 1.44 times as often.<\/p>\r\n<p class=\"oce-p\">Then the number that should hold a service business's attention. Of 714 rendered ads, five (0.7%) came from an advertiser that was also named in the organic recommendation. All five were large national brands, including UrbanStems and Expedia. No small local business appeared to be defending its own organic mention. On local-intent queries the advertiser mix leaned national throughout, led by Expedia, Thumbtack and UrbanStems.<\/p>\r\n<p class=\"oce-p\">Three further splits are worth knowing:<\/p>\r\n<ul class=\"oce-ul\"><li><strong>Commercial phrasing pulls ads.<\/strong> Informational legal questions showed ads on 20.0% of runs. Commercial lawyer-referral requests showed them on 61.8%.<\/li><li><strong>Small markets see fewer.<\/strong> Small towns showed ads on 30.83% of runs, against 54.33% in normal markets.<\/li><li><strong>Auto repair came in lowest<\/strong> at 18.33% ad presence, mainly because fewer ad slots were served.<\/li><\/ul>\r\n<p class=\"oce-p\">Earlier measurements were far lower, which tells you how quickly this is moving.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">Ads are appearing on local queries in ChatGPT, and fast<a class=\"oce-ref\" href=\"#oce-src-10\" title=\"Jump to source 10\">10<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">May to June 2026 (Evertune)<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:19.93%\"><\/span><\/span><span class=\"oce-barval\">13%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">July 23, 2026 (SE Ranking)<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:39.85%\"><\/span><\/span><span class=\"oce-barval\">26%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Late August 2026 (Steady Demand)<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--accent\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">58.25%<\/span><\/div><\/div><p class=\"oce-cnote\">Share of local-intent runs showing a rendered ad. The August figure comes from 1,200 prompts across 10 verticals and six US markets. Of 714 rendered ads, five came from an advertiser also named in the organic recommendation, and all five were national brands.<\/p><\/figure>\r\n<p class=\"oce-p\">The authors are candid about one methodological wobble. Their control group started at 100 runs, and doubling it moved the estimate by 15.5 points and reversed their earlier conclusion. So hold the exact percentages loosely and the pattern firmly.<\/p>\r\n<p class=\"oce-p\">The strategic point survives either way. Paid placement now appears on most local-intent queries in ChatGPT, the advertisers buying it are mostly national aggregators and marketplaces, and the organic recommendation remains the position you can earn with facts.<\/p>\r\n<h3 class=\"oce-h3\">What the buyer on the other side is actually doing<\/h3>\r\n<p class=\"oce-p\">Semrush surveyed 643 US B2B professionals in March and April 2026, keeping 622 valid responses, with 519 AI users forming the analysis base.<\/p>\r\n<p class=\"oce-p\">The adoption figures are no longer arguable. 66% regularly use AI to research products, vendors or solutions, and another 29% do so occasionally. 97% said AI helped them discover a vendor they had not known about, 44% frequently. 92% said AI shaped their shortlist, 45% significantly. 83% said it influenced the final decision, 32% describing that influence as major. 89% expect to lean on AI more for work decisions, and under 1% expect to use it less.<\/p>\r\n<p class=\"oce-p\">Two numbers from this survey should sit on the wall of anyone selling services:<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">51% of respondents used AI to research agencies and service providers<\/span> , the most researched category in the study, ahead of SaaS at 46%. And <strong>84% used AI for purchases of $1,000 or more.<\/strong><\/p>\r\n<p class=\"oce-p\">Then the finding that should rewrite your copy today. Asked what makes them notice a vendor in an AI answer:<\/p>\r\n\r\n<p class=\"oce-p\">Specificity outperforms familiarity by a wide margin, which is good news for anyone who is not the biggest name in their category.<\/p>\r\n<figure class=\"oce-chart\"><figcaption class=\"oce-ctitle\">What makes a buyer notice a vendor in an AI answer<a class=\"oce-ref\" href=\"#oce-src-2\" title=\"Jump to source 2\">2<\/a><\/figcaption><div class=\"oce-bars\"><div class=\"oce-barrow\"><span class=\"oce-barlab\">It closely matches their use case<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:89.29%\"><\/span><\/span><span class=\"oce-barval\">53%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">The description is clear and detailed<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:84.23%\"><\/span><\/span><span class=\"oce-barval\">50%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">Benefits or outcomes are clear<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:64.02%\"><\/span><\/span><span class=\"oce-barval\">38%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">It appears early or is mentioned first<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar\" style=\"width:60.65%\"><\/span><\/span><span class=\"oce-barval\">36%<\/span><\/div><div class=\"oce-barrow\"><span class=\"oce-barlab\">They recognize the brand<\/span><span class=\"oce-bartrack\"><span class=\"oce-bar oce-bar--mute\" style=\"width:11.79%\"><\/span><\/span><span class=\"oce-barval\">7%<\/span><\/div><\/div><p class=\"oce-cnote\">643 US B2B professionals surveyed, 622 valid responses, 519 AI users.<\/p><\/figure>\r\n<p class=\"oce-p\">Their frustrations are equally instructive. 33% said recommendations were too generic for their use case, 28% said they lacked depth or accuracy, 27% said they did not reflect real pricing or contract structures, and 25% said AI missed vendors they knew were relevant. 66% had noticed vendors missing from AI results, 26% of them frequently.<\/p>\r\n<p class=\"oce-p\">The pricing complaint is the argument for publishing numbers, and it comes from buyer behavior. A quarter of buyers are frustrated by the absence of real pricing. If a fixed number is impossible, publish a range and name what moves it.<\/p>\r\n<p class=\"oce-p\">And the click still happens. After seeing a vendor in an AI answer, 71% visit the vendor's website and 63% search for the company by name. 75% still use Google or another search engine for vendor research alongside AI. 41% start with AI and validate through search, 35% start with search and bring in AI, 20% move between both.<\/p>\r\n<p class=\"oce-p\">Your homepage is the page they land on after the assistant names you. It is doing two jobs now: convincing the retrieval layer, then convincing the human it sent.<\/p>\r\n<h3 class=\"oce-h3\">About that 0.28%: the traffic argument, honestly<\/h3>\r\n<p class=\"oce-p\">A study of 74,752 websites by Ahrefs recorded roughly 3.5 million AI chatbot visitors in March 2026, which came to 0.28% of total web traffic. ChatGPT accounted for about 2.7 million of those, roughly ten times any other platform.<\/p>\r\n<p class=\"oce-p\">Under a third of one percent. Easy to dismiss, and plenty of people are dismissing it.<\/p>\r\n<p class=\"oce-p\">Three things complicate that.<\/p>\r\n<p class=\"oce-p\">First, Ahrefs published its own first-party numbers: AI search accounted for 0.5% of their traffic while driving 12.1% of their signups, a conversion rate they calculated at 23 times traditional organic search. Two caveats belong with that figure. Patrick Stox is clear about the limits himself: one company, a 30-day window, free-trial signups, and he expects the rate to fall as volume grows. And that article is older than the traffic data above it, so the two numbers describe different moments. Treat it as directional. Buffer's Simon Heaton separately reported LLM traffic converting at 20.15% against 7.06% for organic search, which points the same way from a different business.<\/p>\r\n<p class=\"oce-p\">Second, and more important: 92% of Semrush's respondents said AI shaped their shortlist. That happens before any click, so referral traffic was never the right place to look for the effect.<\/p>\r\n<p class=\"oce-p\">Third, clicks are becoming rarer while each one carries more weight. In a small study by Kevin Indig covering 48 participants making high-stakes purchases, 64% of Google AI Mode users clicked nothing at all, and the 23% who did click through were mostly confirming a decision they had already made. Small sample, so hold it lightly, but it fits the shortlist finding: by the time someone reaches your site, the comparison may already be over.<\/p>\r\n<h3 class=\"oce-h3\">Four more operational details from the Ahrefs research<\/h3>\r\n<p class=\"oce-p\">These are the specifics that rarely make it into the headlines, and each one is directly actionable.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Freshness is weighted.<\/span> Ahrefs found AI assistants cite content that is 25.7% fresher than what organic search surfaces, with a 13.1% preference for recently updated pages. A price or a service list you last touched in 2024 is a wrong answer being served on repeat.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Each section has to stand alone.<\/span> Ahrefs reports that AI systems consider only the first 30 passages of a page when building embeddings. Front-load the facts, and write sections that make sense when lifted out of context, because that is how they will be read.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Buyers write commands.<\/span> The average ChatGPT prompt runs 42 words, and 75% are phrased as commands rather than questions. Your headings should look like the instruction a buyer types, which is longer and more specific than any keyword you have in a tracking tool.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Roundups are the format assistants trust.<\/span> Among the page types AI systems pull from, best-of lists made up 43.8%. In Ahrefs' cross-site data, best roundups took 7.06% of AI traffic, how-to guides 6.35%, top lists 5.5% and versus comparisons 4.88%. Getting named on a credible best-in-your-city list is leverage you cannot generate from your own domain. Ahrefs also found that branded web mentions showed the strongest correlation with a brand appearing in AI answers (Spearman 0.664), which puts earned mentions and digital PR back on the table as a serious channel.<\/p>\r\n<h3 class=\"oce-h3\">The agents are getting persistent, and that raises the cost of a stale page<\/h3>\r\n<p class=\"oce-p\">Two signals from the first week of November 2026 are worth tracking.<\/p>\r\n<p class=\"oce-p\">Meta launched Spaces on November 6, 2026, giving its Muse agent persistent, topic-focused workspaces that connect third-party apps and refresh in the background, so a price change can surface without anyone asking. On Meta's third-quarter earnings call on October 21, Mark Zuckerberg said more than 10 million people in the US and Canada had set up Muse agents within about a month and a half of the September 8 launch, that roughly a quarter of them used it on any given day, and that shopping was the most common use, with more than a quarter of users shopping through the agent. Spaces arrived one day after Muse Voice.<\/p>\r\n<p class=\"oce-p\">On Product Hunt, November 6 produced two launches in this exact category. Tobira Site Agent led that day's listing, built on the premise that a conventional website sits passive while a buyer's AI assistant arrives to research vendors. Rankfender v3.0 placed third with 127 upvotes, monitoring brand visibility and citations across eight AI assistants including Google AI Overviews.<\/p>\r\n<p class=\"oce-p\">The tooling is arriving because the measurement problem is real. None of it substitutes for having facts on the page worth citing.<\/p>\r\n<h3 class=\"oce-h3\">What to do, in order<\/h3>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">1. Count before you change anything.<\/span> Print your homepage. Highlight every claim a stranger could independently verify: a name, a number, a date, a certification, a specific capability, a named town. Count the highlights. Under twenty, and you have both your diagnosis and your baseline. The sample average was 31.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">2. Run a fact audit with AI rather than by eye.<\/span> The idea of auditing a homepage by fact count belongs to Andy Crestodina and Orbit Media. Here is a prompt you can run as is:<\/p>\r\n<pre class=\"oce-pre\"><code>Audit the homepage copy below and count the specific, independently\r\ncheckable facts it contains.\r\n\r\nContext: a business owner trying to get recommended by AI assistants\r\nwhen a buyer asks for a vendor. Research across 704 websites found\r\nfact count correlates with AI recommendations while word count does not.\r\n\r\nDefinition: a checkable fact is a named client, a result with a number,\r\na credential written as text, a specific capability or piece of\r\nequipment, a named industry or town served, a real price, or a date.\r\nSelf-description is not a fact.\r\n\r\nMy business: [ONE SENTENCE]\r\nMy buyers: [ONE SENTENCE]\r\nMy homepage copy: [PASTE]\r\n\r\nConstraints: use nothing you know about my company from outside this\r\npage, and invent no fact I have not given you.\r\n\r\nReturn:\r\n1. A numbered list of every checkable fact, with a total count.\r\n2. Every self-descriptive adjective to remove, each with one\r\n   replacement sentence built only from facts already on the page.\r\n3. The facts I am missing that a buyer in my category would expect.\r\n\r\nAsk me any questions before you begin.<\/code><\/pre>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">3. Build a fact inventory offline.<\/span> Open a document and list everything true and specific that currently lives in your head or in sales conversations: client names you have written permission to use, outcomes with numbers and dates, certifications and license numbers, equipment and software you actually operate, the industries and towns you serve, real prices. One point from that research is worth repeating: gathering this evidence is the part AI cannot do for you.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">4. Put five or more named clients and real outcomes high on the page.<\/span> Five is where the effect appeared in the data, at 30% against 11% for pages naming none. Get written permission, then name the client with what happened and when. Two named examples carry more weight than ten anonymous testimonials, which showed no measurable effect at all.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">5. Run the subtraction pass.<\/span> Remove every adjective about yourself that a stranger cannot verify: best, leading, world-class, premier, trusted, innovative, seamless, passionate. Pages heavy with that language were recommended 10% of the time against 16% for pages without it. Expect to negotiate with yourself here, and do it in one sitting.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">6. Write out every credential that exists only as an image.<\/span> Badge with text correlated with 19%, badge alone with 10%. While you are there, make sure you have three or more credentials visible, since single certifications showed no measurable effect.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">7. Turn your main headings into the sentence a buyer types.<\/span> Chen and Koudas found question matching dominates everything else. Replace Our Solutions with the actual instruction a buyer gives an assistant at 11pm, in their words and at their length. The average prompt is 42 words.<\/p>\r\n<p class=\"oce-p\"><strong>8. Name six to nine industries or service categories, and your towns explicitly.<\/strong> Six to nine outperformed fewer in the homepage data, and more did not help. For local operators, every town you serve needs to appear in text, not only on a map embed.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">9. Publish real pricing, or a range and what moves it.<\/span> 27% of Semrush's respondents said AI answers do not reflect real pricing or contract structures, which usually means nobody published any. The recommendation-side evidence is neutral. The buyer-side evidence points clearly toward publishing a number.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">10. Complete and maintain your Google Business Profile.<\/span> For local businesses it was the single largest citation source in BrightLocal's data at 28.5%, and AI Mode drew on Google Maps for 79.8% of its local-intent citations. Primary category, every service, service area, hours, photos, and a description written in plain words.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">11. Date-stamp a refresh cycle.<\/span> Ahrefs found a measurable freshness preference. Put a quarterly review on the calendar for prices, service lists, staff credentials and client outcomes.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">12. Measure the rate across repeated checks.<\/span> Same five buyer prompts, four assistants, three runs each, first Monday of the month, written into one log file. Record how many of the sixty checks named you. Without a log you will be guessing in March about whether February was better, and given overlap rates between 20% and 33%, a single check is noise.<\/p>\r\n<h3 class=\"oce-h3\">How to read all of this<\/h3>\r\n<p class=\"oce-p\">Four cautions, because this field is young and the incentives behind the research are not neutral.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Almost every source here sells something adjacent to the finding.<\/span> Ahrefs, Semrush, BrightLocal, Local Falcon, SOCi, Insites, Zarla and Steady Demand all sell tools or services that become more valuable if AI visibility matters, and the homepage study came from an agency. That does not make the data wrong, and several of them publish full methodologies, confidence intervals and raw data, which is more than most marketing research offers. It does mean the findings worth acting on are the ones that replicate across independent samples, and any single vendor's headline number deserves a discount. The academic paper by Chen and Koudas is the one source here with no product attached to the result.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">All of it is correlational.<\/span> The homepage study says so directly. Cheers says so directly. Nobody has yet published a controlled experiment where facts were added to a page and recommendations were measured before and after. Replication across separate samples is what makes the direction credible.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">The absolute numbers conflict, and the dates matter.<\/span> Review-count findings range from a twelvefold gap to a 2.3 times gap depending on the sample. Ad presence in ChatGPT local answers measured roughly 13%, 26% and 58% in three studies run between May and August 2026. Several of the studies carry internal discrepancies in their own reported totals. Where sources disagreed, the ranges in this article are deliberate.<\/p>\r\n<p class=\"oce-p\"><span class=\"oce-lead\">Your market may not behave like the average.<\/span> Every figure above is an average across metros, verticals and engines. Ten service categories in 50 metros tells you very little about your trade in your town, which is the argument for running your own prompts and keeping your own log instead of acting on a national percentage.<\/p>\r\n<p class=\"oce-p\">One number to set against all that caution. A 2026 compilation of local AI search statistics cites Bluehost research finding that only 13% of US small business owners are actively optimizing for AI search. Whatever the true size of this opportunity turns out to be, almost nobody is competing for it yet.<\/p>\r\n<h3 class=\"oce-h3\">Frequently asked questions<\/h3>\r\n<h4 class=\"oce-h4\">Does schema markup or an FAQ section get my business cited?<\/h4>\r\n<p class=\"oce-p\">Less than you have been told. Chen and Koudas, comparing pages within the same website across 158,099 AI answers, found FAQ blocks, structured data, summary boxes and page speed appeared effective only when comparing across different websites, which is a weak and easily confounded comparison. Within a site, question matching explained far more. The homepage study found the same thing from the other direction: FAQ sections with many facts correlated with 28% against 12% for FAQ sections with few, so the facts inside the block were doing the work. Add schema markup where it fits your build. Do not budget for it as a strategy.<\/p>\r\n<h4 class=\"oce-h4\">How many facts does my homepage need?<\/h4>\r\n<p class=\"oce-p\">The homepage data showed recommendation likelihood climbing from roughly 5% at 15 facts to roughly 30% at 60, in a main analysis of 575 companies drawn from 704 collected homepages. The sample average was 31 facts. Treat 60 as a direction of travel. These are correlations from a single study, so no individual fact you add carries a guaranteed return.<\/p>\r\n<p class=\"oce-p\"><strong>Will removing words like leading and best hurt conversion with human visitors?<\/strong><\/p>\r\n<p class=\"oce-p\">No study has tested both audiences at once, so this deserves a test on your own site. The supporting evidence is on the side of specifics: 50% of Semrush's respondents said a clear, detailed description is what makes them notice a vendor, against 7% for brand recognition. Keep your brand voice. Remove the unverifiable claims.<\/p>\r\n<h4 class=\"oce-h4\">AI referral traffic is under 1% of the web. Why does this matter?<\/h4>\r\n<p class=\"oce-p\">Because the decision forms before the click. The Ahrefs study of 74,752 sites found AI chatbots sent 0.28% of total web traffic in March 2026, while its own data showed AI search driving 0.5% of visits and 12.1% of signups. More to the point, 92% of Semrush's respondents said AI had shaped their shortlist, and in Kevin Indig's small study of high-stakes purchases, 64% of Google AI Mode users clicked nothing at all. The shortlist forms whether or not anyone visits.<\/p>\r\n<h4 class=\"oce-h4\">I checked ChatGPT and it recommended me. Am I fine?<\/h4>\r\n<p class=\"oce-p\">One check cannot tell you that. The homepage study found only 4% to 18% of recommended companies appeared in all five runs of the same prompt. BrightLocal measured 20% to 33% overlap between repeat result sets and found 71% to 80% of businesses appeared in half of attempts or fewer. Steady Demand found Gemini returning the same top local business 7.9% of the time, against 90.2% for Google's local pack. Check repeatedly, across engines, and track the rate.<\/p>\r\n<h4 class=\"oce-h4\">My competitor has a worse website and gets recommended anyway. How?<\/h4>\r\n<p class=\"oce-p\">Two likely reasons, both in the research. The homepage study excluded the top quarter of sites by Domain Authority from its main analysis, and that group was recommended 44% of the time, so established authority alone carries recommendations. Separately, Chen and Koudas showed in controlled experiments that brand names alone reordered rankings when the underlying product facts were identical, and that ChatGPT prefers websites it already knows. Authority and familiarity are real. Fact density is the input you can change without waiting years for the other two.<\/p>\r\n<h4 class=\"oce-h4\">Do I need an AI-specific version of my site, or an agent product?<\/h4>\r\n<p class=\"oce-p\">Not before the fundamentals. The fact inventory, the subtraction pass, text-based credentials and a complete Google Business Profile are the foundation, and none of them depend on a new platform. The category is moving, with Tobira Site Agent and Rankfender v3.0 both launching into it on November 6, 2026, and measurement tools are genuinely useful once you have something to measure. They do not substitute for having checkable facts on the page.<\/p>\r\n<h4 class=\"oce-h4\">Does this research apply to local service businesses, or only B2B?<\/h4>\r\n<p class=\"oce-p\">The homepage research and the Semrush survey both covered B2B services and manufacturers, so apply their specific percentages to that context. The local picture comes from BrightLocal's September 2026 study (roughly 200,000 localized prompts, 1.97 million citations, around 1,300 locations) and Steady Demand's August 2026 studies across 50 US metros and 10 local-service verticals. Those two found the same underlying pattern, with a different source mix: your own website supplied 42% of local citations and 93% of unique cited domains, Google Business Profile supplied 28.5%, and everything else was marginal.<\/p>\r\n<h4 class=\"oce-h4\">How many Google reviews do I need?<\/h4>\r\n<p class=\"oce-p\">No honest answer gives you a single number, and anyone quoting a precise threshold is going beyond the evidence. Three separate samples found the same direction and very different magnitudes. Zarla's August 2026 index of 1,851 US local businesses found a median of 444 reviews among businesses the assistants named against 192 among those they skipped. Insites, running 10,000 US local businesses through ChatGPT and Perplexity, found averages of 133.4 against 10.7. What replicates is that review volume separates the two groups while star rating barely does: 4.9 against 4.8 in Zarla's data, and locations ChatGPT recommended in SOCi's index averaged 4.3 stars against a 4.2 average across all locations studied. Build a steady flow of recent reviews and stop chasing the last tenth of a star.<\/p>\r\n<h4 class=\"oce-h4\">Will AI just recommend the big chains over me?<\/h4>\r\n<p class=\"oce-p\">The evidence points the other way. In Local Falcon's June 2026 index of 10,000 US restaurants, independents appeared in Google's AI recommendations 29.1% of the time against 13.9% for chains with five or more locations, even though chains did better on Google Maps. The six chains in that study with more than 100 locations went missing from AI recommendations as often as ordinary restaurants, and sometimes more often. Brand size is not what this layer rewards.<\/p>\r\n<h4 class=\"oce-h4\">Are there ads in AI answers now, and do I need to buy them?<\/h4>\r\n<p class=\"oce-p\">Ads are already there on local queries. In Steady Demand's late-August 2026 test of 1,200 local-intent prompts, ads rendered on 58.25% of runs against 40.50% for prompts with no location. The striking detail is who is buying: of 714 rendered ads, only five came from an advertiser also named in the organic recommendation, and all five were national brands such as Expedia and UrbanStems. Local operators are not yet bidding to defend their own mentions. Whether you should is a budget question for your market. The organic recommendation is the position facts can earn you.<\/p>\r\n<h4 class=\"oce-h4\">Should I be doing this for the assistants or for Google?<\/h4>\r\n<p class=\"oce-p\">Both, and the work overlaps more than the framing suggests. In Zarla's index, every business the assistants named already ranked in Google's top 20 for its trade and city. In BrightLocal's data, Google Business Profile supplied the largest share of local citations at 28.5%, and AI Mode drew on Google Maps for 79.8% of its local-intent citations. Google Maps still named tracked businesses 66% of the time against 32% to 38% for the assistants. Conventional local SEO is the entry requirement, and the fact work is what lifts you inside it.<\/p>\r\n<h3 class=\"oce-h3\">Why this favors smaller operators<\/h3>\r\n<p class=\"oce-p\">For two decades the competitive edge in web copy belonged to whoever could hire the better writer. A human visitor forms an impression in about four seconds, and persuasive adjectives are what shift a person in four seconds, so budget bought credibility.<\/p>\r\n<p class=\"oce-p\">The retrieval layer scores a different thing. It looks for claims with an owner, a number or a date attached, and those come from operating the business rather than from a copy budget. A client list, a set of certifications, an equipment inventory, a service area and a published price belong to whoever did the work to earn them.<\/p>\r\n<p class=\"oce-p\">Across every study in this article the direction holds. Checkable content outperforms polished content. Specificity outperforms familiarity for everyone outside the top tier of domain authority. And roughly a quarter of homepages currently state nothing that matches what their own buyers are asking for, which leaves a wide opening for the companies that close it.<\/p>\r\n<p class=\"oce-p\">Start by counting. Print the homepage, highlight what a stranger could independently verify, and total it. A number under twenty tells you where this quarter's work sits, and it gives you a baseline to measure against in February.<\/p>\r\n<p class=\"oce-p\">Then keep those facts current, and track how often the assistants name you.<\/p>\r\n<div class=\"oce-author\"><div class=\"oce-ava\" role=\"img\" aria-label=\"Scott Kath\"><\/div><div class=\"oce-abody\"><p class=\"oce-aname\">Scott Kath<\/p><p class=\"oce-arole\">President &middot; Oceana PPC Management and SEO Company<\/p><p class=\"oce-abio\">Scott Kath has worked in search marketing for 24 years and has led Oceana PPC Management and SEO Company since 2011, serving businesses of every size across a wide range of industries. His focus now is search everywhere optimization: making one business findable in Google, in the map pack, and inside the AI assistants buyers reach for first. The standard behind that work has not changed in all those years.<\/p><blockquote class=\"oce-aquote\">\u201cWe don\u2019t look for a dollar in savings, we look for 100 pennies.\u201d<\/blockquote><p class=\"oce-abio\">He works directly from Google\u2019s Search Quality Evaluator Guidelines on experience, expertise, authoritativeness and trust, and built Oceana\u2019s Data Driven Sales\u2122 method around measuring what those guidelines reward. Oceana\u2019s work spans technical and local SEO, SEM, conversion rate optimization and web design.<\/p><div class=\"oce-alinks\"><a class=\"oce-alink\" href=\"https:\/\/www.getoceana.com\/marketing\/about-us\/\">About Oceana<\/a><a class=\"oce-alink\" href=\"https:\/\/www.getoceana.com\/marketing\/eeat\/\">EEAT Strategy Guide<\/a><a class=\"oce-alink\" href=\"https:\/\/www.getoceana.com\/marketing\/contact\/\">Contact Scott<\/a><\/div><!-- Social icons: to use the official brand marks, replace the <svg>...<\/svg> inside each link below with the official SVG, then in the CSS set .oce-ico svg to fill:currentColor and remove the stroke rules. Sources: brand.linkedin.com, about.meta.com\/brand\/resources\/facebook, about.meta.com\/brand\/resources\/instagram --><p class=\"oce-amore\"><span class=\"oce-amorelab\">More from Scott<\/span><a class=\"oce-ico\" href=\"https:\/\/www.linkedin.com\/in\/scottkath\" target=\"_blank\" rel=\"noopener\" aria-label=\"LinkedIn\" title=\"LinkedIn\"><!-- LinkedIn icon slot --><svg viewBox=\"0 0 24 24\" aria-hidden=\"true\" focusable=\"false\"><rect x=\"3\" y=\"5\" width=\"18\" height=\"14\" rx=\"2\"\/><circle cx=\"9\" cy=\"11\" r=\"2.2\"\/><path d=\"M5.8 16.2a3.4 3.4 0 0 1 6.4 0M14.5 10h4M14.5 13.5h4\"\/><\/svg><\/a><a class=\"oce-ico\" href=\"https:\/\/scottkath.me\/\" target=\"_blank\" rel=\"noopener\" aria-label=\"scottkath.me\" title=\"scottkath.me\"><svg viewBox=\"0 0 24 24\" aria-hidden=\"true\" focusable=\"false\"><circle cx=\"12\" cy=\"12\" r=\"9\"\/><path d=\"M3 12h18M12 3a15 15 0 0 1 0 18a15 15 0 0 1 0-18\"\/><\/svg><\/a><a class=\"oce-ico\" href=\"https:\/\/scottkath.com\/\" target=\"_blank\" rel=\"noopener\" aria-label=\"scottkath.com\" title=\"scottkath.com\"><svg viewBox=\"0 0 24 24\" aria-hidden=\"true\" focusable=\"false\"><circle cx=\"11\" cy=\"11\" r=\"7\"\/><path d=\"M16.5 16.5 21 21\"\/><\/svg><\/a><a class=\"oce-ico\" href=\"https:\/\/www.facebook.com\/fanalytix\/\" target=\"_blank\" rel=\"noopener\" aria-label=\"Facebook\" title=\"Facebook\"><!-- Facebook icon slot --><svg viewBox=\"0 0 24 24\" aria-hidden=\"true\" focusable=\"false\"><path d=\"M21 11.5a8.5 8.5 0 0 1-8.5 8.5a8.8 8.8 0 0 1-3.7-.8L3 21l1.8-5.3A8.5 8.5 0 1 1 21 11.5Z\"\/><\/svg><\/a><a class=\"oce-ico\" href=\"https:\/\/www.instagram.com\/reallyscottyk\/\" target=\"_blank\" rel=\"noopener\" aria-label=\"Instagram\" title=\"Instagram\"><!-- Instagram icon slot --><svg viewBox=\"0 0 24 24\" aria-hidden=\"true\" focusable=\"false\"><rect x=\"3\" y=\"6\" width=\"18\" height=\"14\" rx=\"3\"\/><circle cx=\"12\" cy=\"13\" r=\"3.6\"\/><path d=\"M8 6l1.4-2.2h5.2L16 6\"\/><\/svg><\/a><\/p><\/div><\/div>\r\n<h3 class=\"oce-h3\">Sources<\/h3>\r\n<ol class=\"oce-srclist\"><li id=\"oce-src-1\"><a href=\"https:\/\/www.orbitmedia.com\/blog\/homepage-facts-ai-recommendations\/\" target=\"_blank\" rel=\"noopener\">What Gets a Brand Recommended by AI? Facts on the Homepage<\/a> by Andy Crestodina, Orbit Media, November 4, 2026<\/li><li id=\"oce-src-2\"><a href=\"https:\/\/www.semrush.com\/blog\/how-ai-shapes-b2b-buying\/\" target=\"_blank\" rel=\"noopener\">How AI tools shape the B2B buying process<\/a> by Margarita Loktionova, Semrush, July 8, 2026<\/li><li id=\"oce-src-3\"><a href=\"https:\/\/arxiv.org\/abs\/2610.15033\" target=\"_blank\" rel=\"noopener\">What Drives AI Citations? Content, Retrieval and Model Knowledge in Generative Engine Optimization<\/a> by Mahe Chen and Nick Koudas, arXiv, October 10, 2026<\/li><li id=\"oce-src-4\"><a href=\"https:\/\/www.brightlocal.com\/research\/local-ai-visibility-study\/\" target=\"_blank\" rel=\"noopener\">Local AI Visibility Study<\/a>, BrightLocal, September 2026<\/li><li id=\"oce-src-5\"><a href=\"https:\/\/www.steadydemand.com\/?p=4957\" target=\"_blank\" rel=\"noopener\">AI Citation Ledger, Grounding Drift, and AI Overviews vs AI Mode<\/a> by Ben Fisher, Steady Demand, August 6 and August 22, 2026<\/li><li id=\"oce-src-6\"><a href=\"https:\/\/ahrefs.com\/blog\/ai-chatbot-traffic\/\" target=\"_blank\" rel=\"noopener\">AI Chatbot Traffic: What It Is, and How to Get More<\/a>, Ahrefs<\/li><li id=\"oce-src-7\"><a href=\"https:\/\/ahrefs.com\/blog\/ai-search-traffic-conversions-ahrefs\/\" target=\"_blank\" rel=\"noopener\">Does AI Search Traffic Convert Better Than Traditional Search?<\/a> by Patrick Stox, Ahrefs<\/li><li id=\"oce-src-8\"><a href=\"https:\/\/www.localfalcon.com\/research\/the-restaurant-ai-visibility-index-how-restaurants-appear-in-ai-search\" target=\"_blank\" rel=\"noopener\">The Restaurant AI Visibility Index<\/a>, Local Falcon, June 12, 2026<\/li><li id=\"oce-src-9\"><a href=\"https:\/\/www.zarla.com\/guides\/ai-visibility-index\" target=\"_blank\" rel=\"noopener\">AI Visibility Index<\/a>, Zarla, August 2026<\/li><li id=\"oce-src-10\"><a href=\"https:\/\/www.steadydemand.com\/?p=4931\" target=\"_blank\" rel=\"noopener\">Ads in ChatGPT local answers<\/a> by Ben Fisher, Steady Demand, August 21, 2026<\/li><li id=\"oce-src-11\"><a href=\"https:\/\/insites.com\/resources\/five-insites\/ai-recommends-businesses-on-review-volume-not-star-rating\/\" target=\"_blank\" rel=\"noopener\">AI recommends businesses on review volume, not star rating<\/a>, Insites, 2026<\/li><li id=\"oce-src-12\"><a href=\"https:\/\/www.cheers.tech\/geo-academy\/reviews-and-geo\" target=\"_blank\" rel=\"noopener\">Do Google Reviews Affect AI Recommendations?<\/a>, Cheers, September 2026<\/li><li id=\"oce-src-13\">SOCi 2026 Local Visibility Index, cited via published summaries of its figures<\/li><li id=\"oce-src-14\"><a href=\"https:\/\/www.bccinteractive.com\/blog\/local-ai-search-statistics\/\" target=\"_blank\" rel=\"noopener\">109 AI Search Statistics For Local And Small Businesses<\/a>, BCC Interactive, 2026, used for the Bluehost and Local Falcon share-of-visibility figures<\/li><li id=\"oce-src-15\"><a href=\"https:\/\/www.producthunt.com\/products\/tobira-ai\" target=\"_blank\" rel=\"noopener\">Tobira Site Agent<\/a>, Product Hunt, November 6, 2026<\/li><li id=\"oce-src-16\"><a href=\"https:\/\/www.producthunt.com\/products\/rankfender\" target=\"_blank\" rel=\"noopener\">Rankfender v3.0<\/a>, Product Hunt, November 6, 2026<\/li><li id=\"oce-src-17\"><a href=\"https:\/\/techcrunch.com\/2026\/11\/06\/meta-introduces-spaces-a-new-way-to-collaborate-with-your-muse-agent\/\" target=\"_blank\" rel=\"noopener\">Meta introduces Spaces, a new way to collaborate with your Muse agent<\/a>, TechCrunch, November 6, 2026<\/li><\/ol>\r\n    <\/div>\r\n  <\/section>\r\n\r\n  <section class=\"oce-cta\">\r\n    <div class=\"oce-cta-inner\">\r\n      <h3>Ready to see your business reach new heights?<\/h3>\r\n      <div class=\"oce-btns\">\r\n        <a class=\"oce-btn\" href=\"https:\/\/www.getoceana.com\/marketing\/website-audits\/\">Website Audits<\/a>\r\n        <a class=\"oce-btn oce-btn-alt\" href=\"https:\/\/www.getoceana.com\/marketing\/contact\/\">Contact Us<\/a>\r\n      <\/div>\r\n    <\/div>\r\n  <\/section>\r\n\r\n<\/div>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"<p>Why AI Recommends Your Competitors and Skips You The 2026 research on this now covers 704 homepages, 158,099 AI answers and several million local citations across a dozen separate studies, and it points one way. The assistant is counting what it can verify on your page, and most homepages give it very little to count&#8230;.<\/p>\n","protected":false},"author":4,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_monsterinsights_skip_tracking":false,"_kadence_starter_templates_imported_post":false,"_kad_post_transparent":"","_kad_post_title":"hide","_kad_post_layout":"fullwidth","_kad_post_sidebar_id":"","_kad_post_content_style":"unboxed","_kad_post_vertical_padding":"hide","_kad_post_feature":"hide","_kad_post_feature_position":"","_kad_post_header":false,"_kad_post_footer":false,"_kad_post_classname":"","_jetpack_hide_featured_image":false,"_jetpack_featured_image_duplicate_dismissed":0,"footnotes":""},"class_list":["post-3902","page","type-page","status-publish","hentry"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.8 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Why AI Recommends Your Competitors and Skips You - Oceana by Scott Kath<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Why AI Recommends Your Competitors and Skips You - Oceana by Scott Kath\" \/>\n<meta property=\"og:description\" content=\"Why AI Recommends Your Competitors and Skips You The 2026 research on this now covers 704 homepages, 158,099 AI answers and several million local citations across a dozen separate studies, and it points one way. The assistant is counting what it can verify on your page, and most homepages give it very little to count....\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/\" \/>\n<meta property=\"og:site_name\" content=\"Oceana by Scott Kath\" \/>\n<meta property=\"article:modified_time\" content=\"2026-11-12T20:14:17+00:00\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data1\" content=\"34 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/why-ai-recommends-your-competitors-and-skips-you\\\/\",\"url\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/why-ai-recommends-your-competitors-and-skips-you\\\/\",\"name\":\"Why AI Recommends Your Competitors and Skips You - Oceana by Scott Kath\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/#website\"},\"datePublished\":\"2026-11-12T02:47:47+00:00\",\"dateModified\":\"2026-11-12T20:14:17+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/why-ai-recommends-your-competitors-and-skips-you\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/why-ai-recommends-your-competitors-and-skips-you\\\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/why-ai-recommends-your-competitors-and-skips-you\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Why AI Recommends Your Competitors and Skips You\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/#website\",\"url\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/\",\"name\":\"Oceana by Scott Kath\",\"description\":\"Data Driven Sales\",\"publisher\":{\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/#organization\",\"name\":\"Oceana by Scott Kath\",\"url\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i0.wp.com\\\/www.getoceana.com\\\/wp-content\\\/uploads\\\/2022\\\/12\\\/cropped-cropped-Oceana_Logo-removebg-preview.png?fit=306%2C123&ssl=1\",\"contentUrl\":\"https:\\\/\\\/i0.wp.com\\\/www.getoceana.com\\\/wp-content\\\/uploads\\\/2022\\\/12\\\/cropped-cropped-Oceana_Logo-removebg-preview.png?fit=306%2C123&ssl=1\",\"width\":306,\"height\":123,\"caption\":\"Oceana by Scott Kath\"},\"image\":{\"@id\":\"https:\\\/\\\/www.getoceana.com\\\/marketing\\\/#\\\/schema\\\/logo\\\/image\\\/\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Why AI Recommends Your Competitors and Skips You - Oceana by Scott Kath","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/","og_locale":"en_US","og_type":"article","og_title":"Why AI Recommends Your Competitors and Skips You - Oceana by Scott Kath","og_description":"Why AI Recommends Your Competitors and Skips You The 2026 research on this now covers 704 homepages, 158,099 AI answers and several million local citations across a dozen separate studies, and it points one way. The assistant is counting what it can verify on your page, and most homepages give it very little to count....","og_url":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/","og_site_name":"Oceana by Scott Kath","article_modified_time":"2026-11-12T20:14:17+00:00","twitter_card":"summary_large_image","twitter_misc":{"Est. reading time":"34 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/","url":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/","name":"Why AI Recommends Your Competitors and Skips You - Oceana by Scott Kath","isPartOf":{"@id":"https:\/\/www.getoceana.com\/marketing\/#website"},"datePublished":"2026-11-12T02:47:47+00:00","dateModified":"2026-11-12T20:14:17+00:00","breadcrumb":{"@id":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.getoceana.com\/marketing\/why-ai-recommends-your-competitors-and-skips-you\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.getoceana.com\/marketing\/"},{"@type":"ListItem","position":2,"name":"Why AI Recommends Your Competitors and Skips You"}]},{"@type":"WebSite","@id":"https:\/\/www.getoceana.com\/marketing\/#website","url":"https:\/\/www.getoceana.com\/marketing\/","name":"Oceana by Scott Kath","description":"Data Driven Sales","publisher":{"@id":"https:\/\/www.getoceana.com\/marketing\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.getoceana.com\/marketing\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.getoceana.com\/marketing\/#organization","name":"Oceana by Scott Kath","url":"https:\/\/www.getoceana.com\/marketing\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.getoceana.com\/marketing\/#\/schema\/logo\/image\/","url":"https:\/\/i0.wp.com\/www.getoceana.com\/wp-content\/uploads\/2022\/12\/cropped-cropped-Oceana_Logo-removebg-preview.png?fit=306%2C123&ssl=1","contentUrl":"https:\/\/i0.wp.com\/www.getoceana.com\/wp-content\/uploads\/2022\/12\/cropped-cropped-Oceana_Logo-removebg-preview.png?fit=306%2C123&ssl=1","width":306,"height":123,"caption":"Oceana by Scott Kath"},"image":{"@id":"https:\/\/www.getoceana.com\/marketing\/#\/schema\/logo\/image\/"}}]}},"jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/pages\/3902","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/comments?post=3902"}],"version-history":[{"count":34,"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/pages\/3902\/revisions"}],"predecessor-version":[{"id":3948,"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/pages\/3902\/revisions\/3948"}],"wp:attachment":[{"href":"https:\/\/www.getoceana.com\/marketing\/wp-json\/wp\/v2\/media?parent=3902"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}