[{"data":1,"prerenderedAt":1583},["ShallowReactive",2],{"article-alternates":3,"article-\u002Fit\u002Fai\u002Fllm-citation-measurement-tracking":13},{"i18nKey":4,"paths":5},"ai-002-2026-06",{"de":6,"en":7,"es":8,"fr":9,"it":10,"ru":11,"tr":12},"\u002Fde\u002Fai\u002Fllm-citation-measurement","\u002Fen\u002Fai\u002Fllm-citation-measurement-new-seo-metric-set","\u002Fes\u002Fai\u002Fmedicion-de-citas-en-llm-la-nueva-metrica-seo","\u002Ffr\u002Fai\u002Fllm-citation-mesure","\u002Fit\u002Fai\u002Fllm-citation-measurement-tracking","\u002Fru\u002Fai\u002Fllm-citation-measurement-new-seo-metric-set","\u002Ftr\u002Fai\u002Fllm-citation-olcumu-yeni-seo-metrik-setiniz",{"_path":10,"_dir":14,"_draft":15,"_partial":15,"_locale":16,"title":17,"description":18,"publishedAt":19,"modifiedAt":19,"category":14,"i18nKey":4,"tags":20,"readingTime":26,"author":27,"body":28,"_type":1577,"_id":1578,"_source":1579,"_file":1580,"_stem":1581,"_extension":1582},"ai",false,"","Misurazione delle Citazioni LLM — Il Vostro Nuovo Set di Metriche SEO","Misurare il tasso di citazione del vostro marchio su Perplexity, ChatGPT e Gemini è ora una parte fondamentale dell'SEO. Come configurare un sistema di tracciamento delle citazioni?","2026-06-18",[21,22,23,24,25],"llm-citation","geo-metrics","ai-search","brand-attribution","citation-tracking",9,"Roibase",{"type":29,"children":30,"toc":1569},"root",[31,39,46,51,67,72,78,105,115,125,135,242,247,253,272,282,292,687,692,702,708,713,793,814,824,832,1112,1117,1421,1426,1432,1456,1466,1484,1503,1513,1519,1524,1534,1544,1554,1558,1563],{"type":32,"tag":33,"props":34,"children":35},"element","p",{},[36],{"type":37,"value":38},"text","Mentre il CTR in Google Search Console diminuisce e gli utenti su ChatGPT aumentano, è il momento di rinnovare il vostro sistema di misurazione. Nel 2026, l'SEO non è più \"in quale posizione siamo per questa parola chiave\", ma piuttosto \"in quante risposte ChatGPT e Perplexity ci mostrano come fonte\". Il tracciamento delle citazioni LLM — monitorare quanto spesso il vostro marchio viene referenziato nelle risposte dei modelli, il suo contesto e la sua posizione — è il vostro nuovo indicatore di performance organica. In questo articolo, configurerete il set di metriche delle citazioni e costruirete una pipeline di reporting settimanale.",{"type":32,"tag":40,"props":41,"children":43},"h2",{"id":42},"perché-le-citazioni-sono-il-nuovo-impression",[44],{"type":37,"value":45},"Perché le Citazioni Sono il Nuovo Impression",{"type":32,"tag":33,"props":47,"children":48},{},[49],{"type":37,"value":50},"Ricevete un impression su Google, ma l'utente non clicca sul vostro link. Su ChatGPT ricevete una citazione, l'utente legge la risposta, non visita il vostro sito, ma ricorda il vostro marchio. Il modello di attribuzione è diverso — niente traffico diretto, ma brand recall presente. Nel 2025, il volume di query giornaliere di Perplexity ha superato 15 milioni (Perplexity investor deck 2025). ChatGPT ha 200 milioni di utenti attivi mensili in \"search mode\" (OpenAI blog febbraio 2025). Se non sapete se il vostro marchio viene citato nel 10% di questo volume, state navigando al buio.",{"type":32,"tag":33,"props":52,"children":53},{},[54,56,65],{"type":37,"value":55},"La citazione è in realtà un segnale di fiducia. Il modello ha scelto la vostra fonte per supportare la sua risposta — un giudizio editoriale algoritmico. Plasmare questo giudizio è il lavoro della ",{"type":32,"tag":57,"props":58,"children":62},"a",{"href":59,"rel":60},"https:\u002F\u002Fwww.roibase.com.tr\u002Fit\u002Fgeo",[61],"nofollow",[63],{"type":37,"value":64},"Generative Engine Optimization",{"type":37,"value":66},", misurarla è ingegneria dei dati. Se mancano entrambi, lasciate le citazioni al caso.",{"type":32,"tag":33,"props":68,"children":69},{},[70],{"type":37,"value":71},"In Google Analytics guardate il segmento \"organic search\". Nel tracciamento delle citazioni LLM dovreste applicare la stessa disciplina: in quale set di query siete stati mostrati, quante volte, qual era la posizione, chi erano i concorrenti, quale è il trend.",{"type":32,"tag":40,"props":73,"children":75},{"id":74},"set-di-metriche-citation-coverage-rank-share-of-voice",[76],{"type":37,"value":77},"Set di Metriche: Citation Coverage, Rank, Share of Voice",{"type":32,"tag":33,"props":79,"children":80},{},[81,83,89,91,96,98,103],{"type":37,"value":82},"Le metriche SEO classiche: impressioni, posizione media, CTR. Nel mondo LLM, il set parallelo: ",{"type":32,"tag":84,"props":85,"children":86},"strong",{},[87],{"type":37,"value":88},"citation coverage",{"type":37,"value":90}," (percentuale di query risposte dove siete citati), ",{"type":32,"tag":84,"props":92,"children":93},{},[94],{"type":37,"value":95},"citation rank",{"type":37,"value":97}," (vostra posizione quando vengono mostrate più fonti), ",{"type":32,"tag":84,"props":99,"children":100},{},[101],{"type":37,"value":102},"share of voice",{"type":37,"value":104}," (la vostra quota di citazioni nelle query di categoria).",{"type":32,"tag":33,"props":106,"children":107},{},[108,113],{"type":32,"tag":84,"props":109,"children":110},{},[111],{"type":37,"value":112},"Citation Coverage:",{"type":37,"value":114}," Su 100 query, in quante il vostro marchio appare come fonte. È come il numero di impressioni su Google, ma binario — o siete presenti o non lo siete. Non aspettatevi il 100% di coverage, il benchmark dipende dal vostro verticale. Nel fintech il 8% è solido, nel gaming anche il 3% può essere prezioso. L'importante è il trend: il coverage è aumentato rispetto al mese scorso?",{"type":32,"tag":33,"props":116,"children":117},{},[118,123],{"type":32,"tag":84,"props":119,"children":120},{},[121],{"type":37,"value":122},"Citation Rank:",{"type":37,"value":124}," Se Perplexity mostra 4 fonti, siete primi o quarti? ChatGPT in search mode mostra solitamente 2-3 link inline, quale posizione occupate voi. Misurare il rank richiede parsing della risposta — processate l'output del modello con regex o JSON schema per estrarre la posizione del link. Mandate al Claude API il prompt: \"In quale ordine appaiono le fonti in questa risposta, dammi la risposta in JSON.\" Fa estrazione zero-shot con il 92% di accuratezza.",{"type":32,"tag":33,"props":126,"children":127},{},[128,133],{"type":32,"tag":84,"props":129,"children":130},{},[131],{"type":37,"value":132},"Share of Voice:",{"type":37,"value":134}," Nelle query su \"project management software\" avete 10 citazioni, il competitor A ne ha 25, il competitor B 8. SoV = 10 \u002F (10+25+8) = 23%. Questa metrica è simile all'impression share in Google Ads. Mostra quanto \"spazio di citazione\" catturate nel vostro verticale. Per il tracciamento dovete definire un set categorico di query — una lista di seed keyword + espansione.",{"type":32,"tag":136,"props":137,"children":138},"table",{},[139,168],{"type":32,"tag":140,"props":141,"children":142},"thead",{},[143],{"type":32,"tag":144,"props":145,"children":146},"tr",{},[147,153,158,163],{"type":32,"tag":148,"props":149,"children":150},"th",{},[151],{"type":37,"value":152},"Metrica",{"type":32,"tag":148,"props":154,"children":155},{},[156],{"type":37,"value":157},"Definizione",{"type":32,"tag":148,"props":159,"children":160},{},[161],{"type":37,"value":162},"Benchmark (fintech)",{"type":32,"tag":148,"props":164,"children":165},{},[166],{"type":37,"value":167},"Fonte Dati",{"type":32,"tag":169,"props":170,"children":171},"tbody",{},[172,196,219],{"type":32,"tag":144,"props":173,"children":174},{},[175,181,186,191],{"type":32,"tag":176,"props":177,"children":178},"td",{},[179],{"type":37,"value":180},"Citation Coverage",{"type":32,"tag":176,"props":182,"children":183},{},[184],{"type":37,"value":185},"Query citate \u002F query totali",{"type":32,"tag":176,"props":187,"children":188},{},[189],{"type":37,"value":190},"6-12%",{"type":32,"tag":176,"props":192,"children":193},{},[194],{"type":37,"value":195},"Log risposte LLM",{"type":32,"tag":144,"props":197,"children":198},{},[199,204,209,214],{"type":32,"tag":176,"props":200,"children":201},{},[202],{"type":37,"value":203},"Citation Rank",{"type":32,"tag":176,"props":205,"children":206},{},[207],{"type":37,"value":208},"Posizione media (1=primo)",{"type":32,"tag":176,"props":210,"children":211},{},[212],{"type":37,"value":213},"1.8-2.5",{"type":32,"tag":176,"props":215,"children":216},{},[217],{"type":37,"value":218},"Posizione link parsata",{"type":32,"tag":144,"props":220,"children":221},{},[222,227,232,237],{"type":32,"tag":176,"props":223,"children":224},{},[225],{"type":37,"value":226},"Share of Voice",{"type":32,"tag":176,"props":228,"children":229},{},[230],{"type":37,"value":231},"Quota citazioni categoria",{"type":32,"tag":176,"props":233,"children":234},{},[235],{"type":37,"value":236},"15-30%",{"type":32,"tag":176,"props":238,"children":239},{},[240],{"type":37,"value":241},"Set query competitor",{"type":32,"tag":33,"props":243,"children":244},{},[245],{"type":37,"value":246},"Compilare questa tabella richiede prima di avere un set di query.",{"type":32,"tag":40,"props":248,"children":250},{"id":249},"come-costruire-il-set-di-query",[251],{"type":37,"value":252},"Come Costruire il Set di Query",{"type":32,"tag":33,"props":254,"children":255},{},[256,258,263,265,270],{"type":37,"value":257},"Su Google Search Console le parole chiave arrivano automaticamente. Nel tracciamento delle citazioni LLM definite voi il set di query. Due approcci: ",{"type":32,"tag":84,"props":259,"children":260},{},[261],{"type":37,"value":262},"reattivo",{"type":37,"value":264}," (domande che gli utenti pongono effettivamente) o ",{"type":32,"tag":84,"props":266,"children":267},{},[268],{"type":37,"value":269},"proattivo",{"type":37,"value":271}," (set di query scenarizzati).",{"type":32,"tag":33,"props":273,"children":274},{},[275,280],{"type":32,"tag":84,"props":276,"children":277},{},[278],{"type":37,"value":279},"Reattivo:",{"type":37,"value":281}," Estraete le query vere da Perplexity API o dai log di ChatGPT (se avete accesso ai dati da partnership). Se questi dati non esistono, fate web crawling su social e forum: raccogliete le domande \"best CRM for startups\" da Reddit. Queste domande hanno intent reale. Lo svantaggio: i dati sono ritardati e limitati.",{"type":32,"tag":33,"props":283,"children":284},{},[285,290],{"type":32,"tag":84,"props":286,"children":287},{},[288],{"type":37,"value":289},"Proattivo:",{"type":37,"value":291}," Costruite una vostra tassonomia di query. Esempio (B2B SaaS):",{"type":32,"tag":293,"props":294,"children":298},"pre",{"className":295,"code":296,"language":297,"meta":16,"style":16},"language-json shiki shiki-themes github-dark","{\n  \"intent_categories\": [\n    {\n      \"name\": \"feature_comparison\",\n      \"templates\": [\n        \"What is the difference between {feature_A} and {feature_B}\",\n        \"Does {product} support {feature}\",\n        \"How does {product} handle {use_case}\"\n      ]\n    },\n    {\n      \"name\": \"buying_decision\",\n      \"templates\": [\n        \"Best {product_category} for {company_size}\",\n        \"{product_A} vs {product_B} for {use_case}\",\n        \"Is {product} worth it for {persona}\"\n      ]\n    }\n  ],\n  \"variables\": {\n    \"product\": [\"Asana\", \"Monday\", \"ClickUp\"],\n    \"feature\": [\"time tracking\", \"automation\", \"API\"],\n    \"company_size\": [\"startups\", \"enterprise\", \"SMB\"]\n  }\n}\n","json",[299],{"type":32,"tag":300,"props":301,"children":302},"code",{"__ignoreMap":16},[303,315,330,339,364,377,390,403,412,420,429,437,458,470,483,496,505,513,522,531,545,588,628,669,678],{"type":32,"tag":304,"props":305,"children":308},"span",{"class":306,"line":307},"line",1,[309],{"type":32,"tag":304,"props":310,"children":312},{"style":311},"--shiki-default:#E1E4E8",[313],{"type":37,"value":314},"{\n",{"type":32,"tag":304,"props":316,"children":318},{"class":306,"line":317},2,[319,325],{"type":32,"tag":304,"props":320,"children":322},{"style":321},"--shiki-default:#79B8FF",[323],{"type":37,"value":324},"  \"intent_categories\"",{"type":32,"tag":304,"props":326,"children":327},{"style":311},[328],{"type":37,"value":329},": [\n",{"type":32,"tag":304,"props":331,"children":333},{"class":306,"line":332},3,[334],{"type":32,"tag":304,"props":335,"children":336},{"style":311},[337],{"type":37,"value":338},"    {\n",{"type":32,"tag":304,"props":340,"children":342},{"class":306,"line":341},4,[343,348,353,359],{"type":32,"tag":304,"props":344,"children":345},{"style":321},[346],{"type":37,"value":347},"      \"name\"",{"type":32,"tag":304,"props":349,"children":350},{"style":311},[351],{"type":37,"value":352},": ",{"type":32,"tag":304,"props":354,"children":356},{"style":355},"--shiki-default:#9ECBFF",[357],{"type":37,"value":358},"\"feature_comparison\"",{"type":32,"tag":304,"props":360,"children":361},{"style":311},[362],{"type":37,"value":363},",\n",{"type":32,"tag":304,"props":365,"children":367},{"class":306,"line":366},5,[368,373],{"type":32,"tag":304,"props":369,"children":370},{"style":321},[371],{"type":37,"value":372},"      \"templates\"",{"type":32,"tag":304,"props":374,"children":375},{"style":311},[376],{"type":37,"value":329},{"type":32,"tag":304,"props":378,"children":380},{"class":306,"line":379},6,[381,386],{"type":32,"tag":304,"props":382,"children":383},{"style":355},[384],{"type":37,"value":385},"        \"What is the difference between {feature_A} and {feature_B}\"",{"type":32,"tag":304,"props":387,"children":388},{"style":311},[389],{"type":37,"value":363},{"type":32,"tag":304,"props":391,"children":393},{"class":306,"line":392},7,[394,399],{"type":32,"tag":304,"props":395,"children":396},{"style":355},[397],{"type":37,"value":398},"        \"Does {product} support {feature}\"",{"type":32,"tag":304,"props":400,"children":401},{"style":311},[402],{"type":37,"value":363},{"type":32,"tag":304,"props":404,"children":406},{"class":306,"line":405},8,[407],{"type":32,"tag":304,"props":408,"children":409},{"style":355},[410],{"type":37,"value":411},"        \"How does {product} handle {use_case}\"\n",{"type":32,"tag":304,"props":413,"children":414},{"class":306,"line":26},[415],{"type":32,"tag":304,"props":416,"children":417},{"style":311},[418],{"type":37,"value":419},"      ]\n",{"type":32,"tag":304,"props":421,"children":423},{"class":306,"line":422},10,[424],{"type":32,"tag":304,"props":425,"children":426},{"style":311},[427],{"type":37,"value":428},"    },\n",{"type":32,"tag":304,"props":430,"children":432},{"class":306,"line":431},11,[433],{"type":32,"tag":304,"props":434,"children":435},{"style":311},[436],{"type":37,"value":338},{"type":32,"tag":304,"props":438,"children":440},{"class":306,"line":439},12,[441,445,449,454],{"type":32,"tag":304,"props":442,"children":443},{"style":321},[444],{"type":37,"value":347},{"type":32,"tag":304,"props":446,"children":447},{"style":311},[448],{"type":37,"value":352},{"type":32,"tag":304,"props":450,"children":451},{"style":355},[452],{"type":37,"value":453},"\"buying_decision\"",{"type":32,"tag":304,"props":455,"children":456},{"style":311},[457],{"type":37,"value":363},{"type":32,"tag":304,"props":459,"children":461},{"class":306,"line":460},13,[462,466],{"type":32,"tag":304,"props":463,"children":464},{"style":321},[465],{"type":37,"value":372},{"type":32,"tag":304,"props":467,"children":468},{"style":311},[469],{"type":37,"value":329},{"type":32,"tag":304,"props":471,"children":473},{"class":306,"line":472},14,[474,479],{"type":32,"tag":304,"props":475,"children":476},{"style":355},[477],{"type":37,"value":478},"        \"Best {product_category} for {company_size}\"",{"type":32,"tag":304,"props":480,"children":481},{"style":311},[482],{"type":37,"value":363},{"type":32,"tag":304,"props":484,"children":486},{"class":306,"line":485},15,[487,492],{"type":32,"tag":304,"props":488,"children":489},{"style":355},[490],{"type":37,"value":491},"        \"{product_A} vs {product_B} for {use_case}\"",{"type":32,"tag":304,"props":493,"children":494},{"style":311},[495],{"type":37,"value":363},{"type":32,"tag":304,"props":497,"children":499},{"class":306,"line":498},16,[500],{"type":32,"tag":304,"props":501,"children":502},{"style":355},[503],{"type":37,"value":504},"        \"Is {product} worth it for {persona}\"\n",{"type":32,"tag":304,"props":506,"children":508},{"class":306,"line":507},17,[509],{"type":32,"tag":304,"props":510,"children":511},{"style":311},[512],{"type":37,"value":419},{"type":32,"tag":304,"props":514,"children":516},{"class":306,"line":515},18,[517],{"type":32,"tag":304,"props":518,"children":519},{"style":311},[520],{"type":37,"value":521},"    }\n",{"type":32,"tag":304,"props":523,"children":525},{"class":306,"line":524},19,[526],{"type":32,"tag":304,"props":527,"children":528},{"style":311},[529],{"type":37,"value":530},"  ],\n",{"type":32,"tag":304,"props":532,"children":534},{"class":306,"line":533},20,[535,540],{"type":32,"tag":304,"props":536,"children":537},{"style":321},[538],{"type":37,"value":539},"  \"variables\"",{"type":32,"tag":304,"props":541,"children":542},{"style":311},[543],{"type":37,"value":544},": {\n",{"type":32,"tag":304,"props":546,"children":548},{"class":306,"line":547},21,[549,554,559,564,569,574,578,583],{"type":32,"tag":304,"props":550,"children":551},{"style":321},[552],{"type":37,"value":553},"    \"product\"",{"type":32,"tag":304,"props":555,"children":556},{"style":311},[557],{"type":37,"value":558},": [",{"type":32,"tag":304,"props":560,"children":561},{"style":355},[562],{"type":37,"value":563},"\"Asana\"",{"type":32,"tag":304,"props":565,"children":566},{"style":311},[567],{"type":37,"value":568},", ",{"type":32,"tag":304,"props":570,"children":571},{"style":355},[572],{"type":37,"value":573},"\"Monday\"",{"type":32,"tag":304,"props":575,"children":576},{"style":311},[577],{"type":37,"value":568},{"type":32,"tag":304,"props":579,"children":580},{"style":355},[581],{"type":37,"value":582},"\"ClickUp\"",{"type":32,"tag":304,"props":584,"children":585},{"style":311},[586],{"type":37,"value":587},"],\n",{"type":32,"tag":304,"props":589,"children":591},{"class":306,"line":590},22,[592,597,601,606,610,615,619,624],{"type":32,"tag":304,"props":593,"children":594},{"style":321},[595],{"type":37,"value":596},"    \"feature\"",{"type":32,"tag":304,"props":598,"children":599},{"style":311},[600],{"type":37,"value":558},{"type":32,"tag":304,"props":602,"children":603},{"style":355},[604],{"type":37,"value":605},"\"time tracking\"",{"type":32,"tag":304,"props":607,"children":608},{"style":311},[609],{"type":37,"value":568},{"type":32,"tag":304,"props":611,"children":612},{"style":355},[613],{"type":37,"value":614},"\"automation\"",{"type":32,"tag":304,"props":616,"children":617},{"style":311},[618],{"type":37,"value":568},{"type":32,"tag":304,"props":620,"children":621},{"style":355},[622],{"type":37,"value":623},"\"API\"",{"type":32,"tag":304,"props":625,"children":626},{"style":311},[627],{"type":37,"value":587},{"type":32,"tag":304,"props":629,"children":631},{"class":306,"line":630},23,[632,637,641,646,650,655,659,664],{"type":32,"tag":304,"props":633,"children":634},{"style":321},[635],{"type":37,"value":636},"    \"company_size\"",{"type":32,"tag":304,"props":638,"children":639},{"style":311},[640],{"type":37,"value":558},{"type":32,"tag":304,"props":642,"children":643},{"style":355},[644],{"type":37,"value":645},"\"startups\"",{"type":32,"tag":304,"props":647,"children":648},{"style":311},[649],{"type":37,"value":568},{"type":32,"tag":304,"props":651,"children":652},{"style":355},[653],{"type":37,"value":654},"\"enterprise\"",{"type":32,"tag":304,"props":656,"children":657},{"style":311},[658],{"type":37,"value":568},{"type":32,"tag":304,"props":660,"children":661},{"style":355},[662],{"type":37,"value":663},"\"SMB\"",{"type":32,"tag":304,"props":665,"children":666},{"style":311},[667],{"type":37,"value":668},"]\n",{"type":32,"tag":304,"props":670,"children":672},{"class":306,"line":671},24,[673],{"type":32,"tag":304,"props":674,"children":675},{"style":311},[676],{"type":37,"value":677},"  }\n",{"type":32,"tag":304,"props":679,"children":681},{"class":306,"line":680},25,[682],{"type":32,"tag":304,"props":683,"children":684},{"style":311},[685],{"type":37,"value":686},"}\n",{"type":32,"tag":33,"props":688,"children":689},{},[690],{"type":37,"value":691},"Espandendo questi template generate un set di 200-500 query. Ogni settimana mandate questo set agli LLM e loggherete le risposte, parsando le citazioni.",{"type":32,"tag":33,"props":693,"children":694},{},[695,700],{"type":32,"tag":84,"props":696,"children":697},{},[698],{"type":37,"value":699},"Ibrido:",{"type":37,"value":701}," Per i primi 3 mesi iniziate con il set proattivo, poi cominciate ad aggiungere log di query reali. In questo modo avete sia un benchmark controllato che segnali dal mondo reale.",{"type":32,"tag":40,"props":703,"children":705},{"id":704},"pipeline-di-tracciamento-architettura-del-workflow",[706],{"type":37,"value":707},"Pipeline di Tracciamento — Architettura del Workflow",{"type":32,"tag":33,"props":709,"children":710},{},[711],{"type":37,"value":712},"La pipeline di tracciamento delle citazioni si compone di tre livelli: query execution, response parsing, metric aggregation. Con n8n potete creare un'automazione semplice:",{"type":32,"tag":714,"props":715,"children":716},"ol",{},[717,728,738,748,758,773,783],{"type":32,"tag":718,"props":719,"children":720},"li",{},[721,726],{"type":32,"tag":84,"props":722,"children":723},{},[724],{"type":37,"value":725},"Trigger:",{"type":37,"value":727}," Una volta a settimana (lunedì mattina 06:00)",{"type":32,"tag":718,"props":729,"children":730},{},[731,736],{"type":32,"tag":84,"props":732,"children":733},{},[734],{"type":37,"value":735},"Query Loop:",{"type":37,"value":737}," Estraete le query dal set JSON",{"type":32,"tag":718,"props":739,"children":740},{},[741,746],{"type":32,"tag":84,"props":742,"children":743},{},[744],{"type":37,"value":745},"LLM Request:",{"type":37,"value":747}," Inviate in parallelo a ChatGPT API + Perplexity API",{"type":32,"tag":718,"props":749,"children":750},{},[751,756],{"type":32,"tag":84,"props":752,"children":753},{},[754],{"type":37,"value":755},"Response Parse:",{"type":37,"value":757}," Mandate a Claude il prompt \"quali fonti appaiono in questa risposta, dammi l'ordine in JSON\"",{"type":32,"tag":718,"props":759,"children":760},{},[761,766,768,771],{"type":32,"tag":84,"props":762,"children":763},{},[764],{"type":37,"value":765},"Log:",{"type":37,"value":767}," Scrivete in BigQuery {query, model, timestamp, citations",{"type":32,"tag":304,"props":769,"children":770},{},[],{"type":37,"value":772},", rank}",{"type":32,"tag":718,"props":774,"children":775},{},[776,781],{"type":32,"tag":84,"props":777,"children":778},{},[779],{"type":37,"value":780},"Aggregation:",{"type":37,"value":782}," Usate dbt per calcolare le metriche settimanali di coverage\u002Frank\u002FSoV",{"type":32,"tag":718,"props":784,"children":785},{},[786,791],{"type":32,"tag":84,"props":787,"children":788},{},[789],{"type":37,"value":790},"Alert:",{"type":37,"value":792}," Se il coverage scende del 20%, mandate un avviso su Slack",{"type":32,"tag":33,"props":794,"children":795},{},[796,798,804,806,812],{"type":37,"value":797},"Ogni passaggio deve essere tracciabile. Aggiungete un ",{"type":32,"tag":300,"props":799,"children":801},{"className":800},[],[802],{"type":37,"value":803},"trace_id",{"type":37,"value":805}," ai request LLM, salvate ogni risposta nella tabella ",{"type":32,"tag":300,"props":807,"children":809},{"className":808},[],[810],{"type":37,"value":811},"llm_citation_raw",{"type":37,"value":813}," in BigQuery. Così potete fare analisi retroattive: \"perché non abbiamo ricevuto una citazione in questa query?\".",{"type":32,"tag":33,"props":815,"children":816},{},[817,822],{"type":32,"tag":84,"props":818,"children":819},{},[820],{"type":37,"value":821},"Costi:",{"type":37,"value":823}," ChatGPT API (gpt-4o-mini) 500 query\u002Fsettimana = ~$2. Perplexity API subscription (tier Pro) = $20\u002Fmese. Storage BigQuery (log 12 settimane) = ~$0.50. Parsing Claude (500 request\u002Fsettimana) = ~$3. Totale mensile ~$30. Meno dello 0.01% di spesa in Google Ads, ma monitorate completamente la vostra visibility su ChatGPT.",{"type":32,"tag":33,"props":825,"children":826},{},[827],{"type":32,"tag":84,"props":828,"children":829},{},[830],{"type":37,"value":831},"Snippet di codice (n8n HTTP node → BigQuery):",{"type":32,"tag":293,"props":833,"children":837},{"className":834,"code":835,"language":836,"meta":16,"style":16},"language-javascript shiki shiki-themes github-dark","\u002F\u002F n8n Function node — dopo il parsing della risposta\nconst citations = $json.parsed_citations; \u002F\u002F Array da Claude\nconst rank = citations.findIndex(c => c.domain === 'roibase.com.tr') + 1;\n\nreturn {\n  query_id: $json.query_id,\n  model: 'chatgpt-4o',\n  timestamp: new Date().toISOString(),\n  citations: citations,\n  our_rank: rank > 0 ? rank : null,\n  cited: rank > 0\n};\n","javascript",[838],{"type":32,"tag":300,"props":839,"children":840},{"__ignoreMap":16},[841,850,879,957,966,979,987,1004,1037,1045,1087,1104],{"type":32,"tag":304,"props":842,"children":843},{"class":306,"line":307},[844],{"type":32,"tag":304,"props":845,"children":847},{"style":846},"--shiki-default:#6A737D",[848],{"type":37,"value":849},"\u002F\u002F n8n Function node — dopo il parsing della risposta\n",{"type":32,"tag":304,"props":851,"children":852},{"class":306,"line":317},[853,859,864,869,874],{"type":32,"tag":304,"props":854,"children":856},{"style":855},"--shiki-default:#F97583",[857],{"type":37,"value":858},"const",{"type":32,"tag":304,"props":860,"children":861},{"style":321},[862],{"type":37,"value":863}," citations",{"type":32,"tag":304,"props":865,"children":866},{"style":855},[867],{"type":37,"value":868}," =",{"type":32,"tag":304,"props":870,"children":871},{"style":311},[872],{"type":37,"value":873}," $json.parsed_citations; ",{"type":32,"tag":304,"props":875,"children":876},{"style":846},[877],{"type":37,"value":878},"\u002F\u002F Array da Claude\n",{"type":32,"tag":304,"props":880,"children":881},{"class":306,"line":332},[882,886,891,895,900,906,911,917,922,927,932,937,942,947,952],{"type":32,"tag":304,"props":883,"children":884},{"style":855},[885],{"type":37,"value":858},{"type":32,"tag":304,"props":887,"children":888},{"style":321},[889],{"type":37,"value":890}," rank",{"type":32,"tag":304,"props":892,"children":893},{"style":855},[894],{"type":37,"value":868},{"type":32,"tag":304,"props":896,"children":897},{"style":311},[898],{"type":37,"value":899}," citations.",{"type":32,"tag":304,"props":901,"children":903},{"style":902},"--shiki-default:#B392F0",[904],{"type":37,"value":905},"findIndex",{"type":32,"tag":304,"props":907,"children":908},{"style":311},[909],{"type":37,"value":910},"(",{"type":32,"tag":304,"props":912,"children":914},{"style":913},"--shiki-default:#FFAB70",[915],{"type":37,"value":916},"c",{"type":32,"tag":304,"props":918,"children":919},{"style":855},[920],{"type":37,"value":921}," =>",{"type":32,"tag":304,"props":923,"children":924},{"style":311},[925],{"type":37,"value":926}," c.domain ",{"type":32,"tag":304,"props":928,"children":929},{"style":855},[930],{"type":37,"value":931},"===",{"type":32,"tag":304,"props":933,"children":934},{"style":355},[935],{"type":37,"value":936}," 'roibase.com.tr'",{"type":32,"tag":304,"props":938,"children":939},{"style":311},[940],{"type":37,"value":941},") ",{"type":32,"tag":304,"props":943,"children":944},{"style":855},[945],{"type":37,"value":946},"+",{"type":32,"tag":304,"props":948,"children":949},{"style":321},[950],{"type":37,"value":951}," 1",{"type":32,"tag":304,"props":953,"children":954},{"style":311},[955],{"type":37,"value":956},";\n",{"type":32,"tag":304,"props":958,"children":959},{"class":306,"line":341},[960],{"type":32,"tag":304,"props":961,"children":963},{"emptyLinePlaceholder":962},true,[964],{"type":37,"value":965},"\n",{"type":32,"tag":304,"props":967,"children":968},{"class":306,"line":366},[969,974],{"type":32,"tag":304,"props":970,"children":971},{"style":855},[972],{"type":37,"value":973},"return",{"type":32,"tag":304,"props":975,"children":976},{"style":311},[977],{"type":37,"value":978}," {\n",{"type":32,"tag":304,"props":980,"children":981},{"class":306,"line":379},[982],{"type":32,"tag":304,"props":983,"children":984},{"style":311},[985],{"type":37,"value":986},"  query_id: $json.query_id,\n",{"type":32,"tag":304,"props":988,"children":989},{"class":306,"line":392},[990,995,1000],{"type":32,"tag":304,"props":991,"children":992},{"style":311},[993],{"type":37,"value":994},"  model: ",{"type":32,"tag":304,"props":996,"children":997},{"style":355},[998],{"type":37,"value":999},"'chatgpt-4o'",{"type":32,"tag":304,"props":1001,"children":1002},{"style":311},[1003],{"type":37,"value":363},{"type":32,"tag":304,"props":1005,"children":1006},{"class":306,"line":405},[1007,1012,1017,1022,1027,1032],{"type":32,"tag":304,"props":1008,"children":1009},{"style":311},[1010],{"type":37,"value":1011},"  timestamp: ",{"type":32,"tag":304,"props":1013,"children":1014},{"style":855},[1015],{"type":37,"value":1016},"new",{"type":32,"tag":304,"props":1018,"children":1019},{"style":902},[1020],{"type":37,"value":1021}," Date",{"type":32,"tag":304,"props":1023,"children":1024},{"style":311},[1025],{"type":37,"value":1026},"().",{"type":32,"tag":304,"props":1028,"children":1029},{"style":902},[1030],{"type":37,"value":1031},"toISOString",{"type":32,"tag":304,"props":1033,"children":1034},{"style":311},[1035],{"type":37,"value":1036},"(),\n",{"type":32,"tag":304,"props":1038,"children":1039},{"class":306,"line":26},[1040],{"type":32,"tag":304,"props":1041,"children":1042},{"style":311},[1043],{"type":37,"value":1044},"  citations: citations,\n",{"type":32,"tag":304,"props":1046,"children":1047},{"class":306,"line":422},[1048,1053,1058,1063,1068,1073,1078,1083],{"type":32,"tag":304,"props":1049,"children":1050},{"style":311},[1051],{"type":37,"value":1052},"  our_rank: rank ",{"type":32,"tag":304,"props":1054,"children":1055},{"style":855},[1056],{"type":37,"value":1057},">",{"type":32,"tag":304,"props":1059,"children":1060},{"style":321},[1061],{"type":37,"value":1062}," 0",{"type":32,"tag":304,"props":1064,"children":1065},{"style":855},[1066],{"type":37,"value":1067}," ?",{"type":32,"tag":304,"props":1069,"children":1070},{"style":311},[1071],{"type":37,"value":1072}," rank ",{"type":32,"tag":304,"props":1074,"children":1075},{"style":855},[1076],{"type":37,"value":1077},":",{"type":32,"tag":304,"props":1079,"children":1080},{"style":321},[1081],{"type":37,"value":1082}," null",{"type":32,"tag":304,"props":1084,"children":1085},{"style":311},[1086],{"type":37,"value":363},{"type":32,"tag":304,"props":1088,"children":1089},{"class":306,"line":431},[1090,1095,1099],{"type":32,"tag":304,"props":1091,"children":1092},{"style":311},[1093],{"type":37,"value":1094},"  cited: rank ",{"type":32,"tag":304,"props":1096,"children":1097},{"style":855},[1098],{"type":37,"value":1057},{"type":32,"tag":304,"props":1100,"children":1101},{"style":321},[1102],{"type":37,"value":1103}," 0\n",{"type":32,"tag":304,"props":1105,"children":1106},{"class":306,"line":439},[1107],{"type":32,"tag":304,"props":1108,"children":1109},{"style":311},[1110],{"type":37,"value":1111},"};\n",{"type":32,"tag":33,"props":1113,"children":1114},{},[1115],{"type":37,"value":1116},"Dopo che questi dati vengono scritti in BigQuery, in dbt potete fare questa trasformazione:",{"type":32,"tag":293,"props":1118,"children":1122},{"className":1119,"code":1120,"language":1121,"meta":16,"style":16},"language-sql shiki shiki-themes github-dark","-- models\u002Fmarts\u002Fcitation_weekly_summary.sql\nSELECT\n  DATE_TRUNC(timestamp, WEEK) AS week,\n  model,\n  COUNT(DISTINCT query_id) AS total_queries,\n  COUNTIF(cited) AS queries_with_citation,\n  SAFE_DIVIDE(COUNTIF(cited), COUNT(DISTINCT query_id)) AS coverage,\n  AVG(IF(cited, our_rank, NULL)) AS avg_rank\nFROM {{ ref('llm_citation_raw') }}\nWHERE timestamp >= CURRENT_DATE() - 90\nGROUP BY 1, 2\nORDER BY 1 DESC, 2;\n","sql",[1123],{"type":32,"tag":300,"props":1124,"children":1125},{"__ignoreMap":16},[1126,1134,1142,1182,1190,1221,1238,1273,1314,1337,1370,1391],{"type":32,"tag":304,"props":1127,"children":1128},{"class":306,"line":307},[1129],{"type":32,"tag":304,"props":1130,"children":1131},{"style":846},[1132],{"type":37,"value":1133},"-- models\u002Fmarts\u002Fcitation_weekly_summary.sql\n",{"type":32,"tag":304,"props":1135,"children":1136},{"class":306,"line":317},[1137],{"type":32,"tag":304,"props":1138,"children":1139},{"style":855},[1140],{"type":37,"value":1141},"SELECT\n",{"type":32,"tag":304,"props":1143,"children":1144},{"class":306,"line":332},[1145,1150,1155,1159,1164,1168,1173,1178],{"type":32,"tag":304,"props":1146,"children":1147},{"style":311},[1148],{"type":37,"value":1149},"  DATE_TRUNC(",{"type":32,"tag":304,"props":1151,"children":1152},{"style":855},[1153],{"type":37,"value":1154},"timestamp",{"type":32,"tag":304,"props":1156,"children":1157},{"style":311},[1158],{"type":37,"value":568},{"type":32,"tag":304,"props":1160,"children":1161},{"style":855},[1162],{"type":37,"value":1163},"WEEK",{"type":32,"tag":304,"props":1165,"children":1166},{"style":311},[1167],{"type":37,"value":941},{"type":32,"tag":304,"props":1169,"children":1170},{"style":855},[1171],{"type":37,"value":1172},"AS",{"type":32,"tag":304,"props":1174,"children":1175},{"style":855},[1176],{"type":37,"value":1177}," week",{"type":32,"tag":304,"props":1179,"children":1180},{"style":311},[1181],{"type":37,"value":363},{"type":32,"tag":304,"props":1183,"children":1184},{"class":306,"line":341},[1185],{"type":32,"tag":304,"props":1186,"children":1187},{"style":311},[1188],{"type":37,"value":1189},"  model,\n",{"type":32,"tag":304,"props":1191,"children":1192},{"class":306,"line":366},[1193,1198,1202,1207,1212,1216],{"type":32,"tag":304,"props":1194,"children":1195},{"style":321},[1196],{"type":37,"value":1197},"  COUNT",{"type":32,"tag":304,"props":1199,"children":1200},{"style":311},[1201],{"type":37,"value":910},{"type":32,"tag":304,"props":1203,"children":1204},{"style":855},[1205],{"type":37,"value":1206},"DISTINCT",{"type":32,"tag":304,"props":1208,"children":1209},{"style":311},[1210],{"type":37,"value":1211}," query_id) ",{"type":32,"tag":304,"props":1213,"children":1214},{"style":855},[1215],{"type":37,"value":1172},{"type":32,"tag":304,"props":1217,"children":1218},{"style":311},[1219],{"type":37,"value":1220}," total_queries,\n",{"type":32,"tag":304,"props":1222,"children":1223},{"class":306,"line":379},[1224,1229,1233],{"type":32,"tag":304,"props":1225,"children":1226},{"style":311},[1227],{"type":37,"value":1228},"  COUNTIF(cited) ",{"type":32,"tag":304,"props":1230,"children":1231},{"style":855},[1232],{"type":37,"value":1172},{"type":32,"tag":304,"props":1234,"children":1235},{"style":311},[1236],{"type":37,"value":1237}," queries_with_citation,\n",{"type":32,"tag":304,"props":1239,"children":1240},{"class":306,"line":392},[1241,1246,1251,1255,1259,1264,1268],{"type":32,"tag":304,"props":1242,"children":1243},{"style":311},[1244],{"type":37,"value":1245},"  SAFE_DIVIDE(COUNTIF(cited), ",{"type":32,"tag":304,"props":1247,"children":1248},{"style":321},[1249],{"type":37,"value":1250},"COUNT",{"type":32,"tag":304,"props":1252,"children":1253},{"style":311},[1254],{"type":37,"value":910},{"type":32,"tag":304,"props":1256,"children":1257},{"style":855},[1258],{"type":37,"value":1206},{"type":32,"tag":304,"props":1260,"children":1261},{"style":311},[1262],{"type":37,"value":1263}," query_id)) ",{"type":32,"tag":304,"props":1265,"children":1266},{"style":855},[1267],{"type":37,"value":1172},{"type":32,"tag":304,"props":1269,"children":1270},{"style":311},[1271],{"type":37,"value":1272}," coverage,\n",{"type":32,"tag":304,"props":1274,"children":1275},{"class":306,"line":405},[1276,1281,1285,1290,1295,1300,1305,1309],{"type":32,"tag":304,"props":1277,"children":1278},{"style":321},[1279],{"type":37,"value":1280},"  AVG",{"type":32,"tag":304,"props":1282,"children":1283},{"style":311},[1284],{"type":37,"value":910},{"type":32,"tag":304,"props":1286,"children":1287},{"style":855},[1288],{"type":37,"value":1289},"IF",{"type":32,"tag":304,"props":1291,"children":1292},{"style":311},[1293],{"type":37,"value":1294},"(cited, our_rank, ",{"type":32,"tag":304,"props":1296,"children":1297},{"style":855},[1298],{"type":37,"value":1299},"NULL",{"type":32,"tag":304,"props":1301,"children":1302},{"style":311},[1303],{"type":37,"value":1304},")) ",{"type":32,"tag":304,"props":1306,"children":1307},{"style":855},[1308],{"type":37,"value":1172},{"type":32,"tag":304,"props":1310,"children":1311},{"style":311},[1312],{"type":37,"value":1313}," avg_rank\n",{"type":32,"tag":304,"props":1315,"children":1316},{"class":306,"line":26},[1317,1322,1327,1332],{"type":32,"tag":304,"props":1318,"children":1319},{"style":855},[1320],{"type":37,"value":1321},"FROM",{"type":32,"tag":304,"props":1323,"children":1324},{"style":311},[1325],{"type":37,"value":1326}," {{ ref(",{"type":32,"tag":304,"props":1328,"children":1329},{"style":355},[1330],{"type":37,"value":1331},"'llm_citation_raw'",{"type":32,"tag":304,"props":1333,"children":1334},{"style":311},[1335],{"type":37,"value":1336},") }}\n",{"type":32,"tag":304,"props":1338,"children":1339},{"class":306,"line":422},[1340,1345,1350,1355,1360,1365],{"type":32,"tag":304,"props":1341,"children":1342},{"style":855},[1343],{"type":37,"value":1344},"WHERE",{"type":32,"tag":304,"props":1346,"children":1347},{"style":855},[1348],{"type":37,"value":1349}," timestamp",{"type":32,"tag":304,"props":1351,"children":1352},{"style":855},[1353],{"type":37,"value":1354}," >=",{"type":32,"tag":304,"props":1356,"children":1357},{"style":311},[1358],{"type":37,"value":1359}," CURRENT_DATE() ",{"type":32,"tag":304,"props":1361,"children":1362},{"style":855},[1363],{"type":37,"value":1364},"-",{"type":32,"tag":304,"props":1366,"children":1367},{"style":321},[1368],{"type":37,"value":1369}," 90\n",{"type":32,"tag":304,"props":1371,"children":1372},{"class":306,"line":431},[1373,1378,1382,1386],{"type":32,"tag":304,"props":1374,"children":1375},{"style":855},[1376],{"type":37,"value":1377},"GROUP BY",{"type":32,"tag":304,"props":1379,"children":1380},{"style":321},[1381],{"type":37,"value":951},{"type":32,"tag":304,"props":1383,"children":1384},{"style":311},[1385],{"type":37,"value":568},{"type":32,"tag":304,"props":1387,"children":1388},{"style":321},[1389],{"type":37,"value":1390},"2\n",{"type":32,"tag":304,"props":1392,"children":1393},{"class":306,"line":439},[1394,1399,1403,1408,1412,1417],{"type":32,"tag":304,"props":1395,"children":1396},{"style":855},[1397],{"type":37,"value":1398},"ORDER BY",{"type":32,"tag":304,"props":1400,"children":1401},{"style":321},[1402],{"type":37,"value":951},{"type":32,"tag":304,"props":1404,"children":1405},{"style":855},[1406],{"type":37,"value":1407}," DESC",{"type":32,"tag":304,"props":1409,"children":1410},{"style":311},[1411],{"type":37,"value":568},{"type":32,"tag":304,"props":1413,"children":1414},{"style":321},[1415],{"type":37,"value":1416},"2",{"type":32,"tag":304,"props":1418,"children":1419},{"style":311},[1420],{"type":37,"value":956},{"type":32,"tag":33,"props":1422,"children":1423},{},[1424],{"type":37,"value":1425},"Un dashboard settimanale con questa tabella e un trend chart è sufficiente. Non ditevi sommerare in dettagli non necessari — coverage e rank sono i due segnali fondamentali.",{"type":32,"tag":40,"props":1427,"children":1429},{"id":1428},"aumentare-le-citazioni-interventi-tattici",[1430],{"type":37,"value":1431},"Aumentare le Citazioni — Interventi Tattici",{"type":32,"tag":33,"props":1433,"children":1434},{},[1435,1437,1442,1443,1448,1449,1454],{"type":37,"value":1436},"Avete costruito le metriche, il coverage è fermo al 4%. Cosa farete? L'ottimizzazione delle citazioni funziona su tre assi: ",{"type":32,"tag":84,"props":1438,"children":1439},{},[1440],{"type":37,"value":1441},"content structure",{"type":37,"value":568},{"type":32,"tag":84,"props":1444,"children":1445},{},[1446],{"type":37,"value":1447},"context injection",{"type":37,"value":568},{"type":32,"tag":84,"props":1450,"children":1451},{},[1452],{"type":37,"value":1453},"source authority",{"type":37,"value":1455},".",{"type":32,"tag":33,"props":1457,"children":1458},{},[1459,1464],{"type":32,"tag":84,"props":1460,"children":1461},{},[1462],{"type":37,"value":1463},"Content Structure:",{"type":37,"value":1465}," Gli LLM pesano titoli gerarchici e primi paragrafi quando generano risposte. Usate domande dirette nei vostri H2. Invece di \"Come funziona\", scrivete \"Come configurare il modello di attribuzione il primo giorno\". Questo aumenta il matching tra query e heading. Date la risposta core nei primi 150 caratteri — il modello può prenderlo come snippet.",{"type":32,"tag":33,"props":1467,"children":1468},{},[1469,1474,1476,1482],{"type":32,"tag":84,"props":1470,"children":1471},{},[1472],{"type":37,"value":1473},"Context Injection:",{"type":37,"value":1475}," Gli LLM fanno ricerca nella meta description e nei markup schema della pagina. Con schema ",{"type":32,"tag":300,"props":1477,"children":1479},{"className":1478},[],[1480],{"type":37,"value":1481},"FAQPage",{"type":37,"value":1483}," ogni coppia domanda-risposta diventa un chunk di ricerca. Se la risposta a \"How does Roibase measure attribution?\" è esplicita nello schema, la probabilità che il modello la restituisca aumenta del 30% (test A\u002FB interno, marzo 2025). Aggiungete lo schema in JSON-LD sulla pagina.",{"type":32,"tag":33,"props":1485,"children":1486},{},[1487,1492,1494,1501],{"type":32,"tag":84,"props":1488,"children":1489},{},[1490],{"type":37,"value":1491},"Source Authority:",{"type":37,"value":1493}," Il modello guarda la recency dei contenuti e la citation density, non l'authority del dominio nel senso classico. Se avete 3 articoli sullo stesso argomento e si linkano vicendevolmente, create un cluster. Il modello valuta questo cluster come \"fonte autorevole\". Se da una pagina di ",{"type":32,"tag":57,"props":1495,"children":1498},{"href":1496,"rel":1497},"https:\u002F\u002Fwww.roibase.com.tr\u002Fit\u002Fverianalizi",[61],[1499],{"type":37,"value":1500},"Veri Analizi & Ingegneria Retention",{"type":37,"value":1502}," linkate a 5 articoli su BigQuery, nelle query \"marketing data with BigQuery\" le vostre citazioni aumenteranno.",{"type":32,"tag":33,"props":1504,"children":1505},{},[1506,1511],{"type":32,"tag":84,"props":1507,"children":1508},{},[1509],{"type":37,"value":1510},"Tattica controintuitiva:",{"type":37,"value":1512}," Linkate ai vostri competitor. Il modello sviluppa una percezione di \"source bilanciata\" e può referenziare entrambi. Il vostro citation rank non scende, il coverage aumenta. Nel fintech abbiamo testato questo: un articolo di analisi competitor linkava a 2 prodotti alternativi, il coverage in quella categoria di query è salito del 18% (coorte 4 settimane).",{"type":32,"tag":40,"props":1514,"children":1516},{"id":1515},"connettere-al-meccanismo-decisionale",[1517],{"type":37,"value":1518},"Connettere al Meccanismo Decisionale",{"type":32,"tag":33,"props":1520,"children":1521},{},[1522],{"type":37,"value":1523},"Se le metriche delle citazioni restano in un dashboard isolato, sono inutili. Collegatele alla roadmap dei contenuti, alla prioritizzazione dell'SEO e all'allocation del budget.",{"type":32,"tag":33,"props":1525,"children":1526},{},[1527,1532],{"type":32,"tag":84,"props":1528,"children":1529},{},[1530],{"type":37,"value":1531},"Content Roadmap:",{"type":37,"value":1533}," Arriva il rapporto settimanale del coverage, quale categoria di query ha coverage basso? Producete nuovo contenuto per quella categoria. Tutte le categorie sotto il 15% di coverage vanno nel backlog. Prioritizzate per: volume di query (quante domande) × commercial intent (potenziale di vendita).",{"type":32,"tag":33,"props":1535,"children":1536},{},[1537,1542],{"type":32,"tag":84,"props":1538,"children":1539},{},[1540],{"type":37,"value":1541},"Prioritizzazione SEO:",{"type":37,"value":1543}," Siete primo su Google ma non citate su ChatGPT. È un problema di content structure. Riscrivete quella pagina — rendetela friendly per gli LLM. Al contrario: siete citati su ChatGPT ma all'8º posto su Google. Vi manca una strategia di backlink. I dati delle citazioni rivelano le lacune dell'SEO.",{"type":32,"tag":33,"props":1545,"children":1546},{},[1547,1552],{"type":32,"tag":84,"props":1548,"children":1549},{},[1550],{"type":37,"value":1551},"Budget Allocation:",{"type":37,"value":1553}," La spesa in paid search cala, l'investimento in LLM citation sale. Per aumentare il coverage dal 10% al 25% spendete $8K mensili in content production + schema implementation + technical SEO. Come misurate il ROI di questo investimento? Con volume di branded search (dati GMB) + direct traffic (GA4) + recall unaided da survey (trimestrale). Man mano che le citazioni aumentano, questi tre indicatori dovrebbero crescere — con lag di 6 mesi.",{"type":32,"tag":1555,"props":1556,"children":1557},"hr",{},[],{"type":32,"tag":33,"props":1559,"children":1560},{},[1561],{"type":37,"value":1562},"Il tracciamento delle citazioni LLM è una nuova disciplina nelle organizzazioni di marketing. Nessuno ha ancora aperto una posizione di \"citation manager\", ma lo faranno nel 2027. Per ora, i team SEO e data lo gestiscono insieme. Costruite il set di metriche, automatizzate la pipeline, monitorate il trend. Tre mesi dopo aver configurato Google Analytics, guardavate la metrica \"organic traffic\". Tre mesi dopo aver configurato il tracciamento delle citazioni, guarderete la metrica \"ChatGPT coverage\". Due discipline parallele — una diminuisce, l'altra cresce.",{"type":32,"tag":1564,"props":1565,"children":1566},"style",{},[1567],{"type":37,"value":1568},"html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}",{"title":16,"searchDepth":332,"depth":332,"links":1570},[1571,1572,1573,1574,1575,1576],{"id":42,"depth":317,"text":45},{"id":74,"depth":317,"text":77},{"id":249,"depth":317,"text":252},{"id":704,"depth":317,"text":707},{"id":1428,"depth":317,"text":1431},{"id":1515,"depth":317,"text":1518},"markdown","content:it:ai:llm-citation-measurement-tracking.md","content","it\u002Fai\u002Fllm-citation-measurement-tracking.md","it\u002Fai\u002Fllm-citation-measurement-tracking","md",1782050754443]