{
  "kind": "answer",
  "studySlug": "query-translation",
  "slug": "how-aggressive-are-ai-query-rewrites",
  "title": "How aggressive are AI query rewrites?",
  "description": "They are aggressive. Average similarity between prompt and search query is only 25.24%, and 31.85% of pairs count as complete rewrites.",
  "lastUpdated": "2026-01-22",
  "lastTested": "2026-01-22",
  "sourceStudyUrl": "/trakkr-research/query-translation",
  "sourceStudyTitle": "How AI Translates Your Questions",
  "claimIds": [
    "query-translation:similarity",
    "query-translation:complete-rewrite",
    "query-translation:pairs"
  ],
  "relatedSlugs": [
    "answer:does-ai-add-current-year-terms-to-searches",
    "answer:does-ai-turn-vague-prompts-into-lists-and-best-of-queries",
    "fact:average-prompt-to-query-similarity-is-only-twenty-five-percent",
    "tracker:query-transformation-severity-tracker"
  ],
  "methodologySummary": "Built from 11,521 captured prompt-to-query pairs observed in OpenAI web search calls, with 100% search-query coverage in the sampled dataset.",
  "limitations": [
    "This study reflects captured search-query transformations rather than every hidden retrieval step a model may perform.",
    "Some inserted entities reflect category heuristics or market leaders rather than final recommendation outcomes.",
    "The underlying dataset is concentrated in commercial and research prompts rather than all possible conversational traffic."
  ],
  "keywords": [
    "query translation",
    "AI search rewrites",
    "prompt rewriting",
    "AI retrieval behavior",
    "aggressive query rewrites",
    "AI rewrite rate"
  ],
  "schemaHints": {
    "pageType": "Article",
    "includeDataset": true
  },
  "question": "How aggressive are AI query rewrites?",
  "directAnswer": "Mostly, they are aggressive. Average Jaccard similarity between prompt and search query is only 25.24%, and 31.85% of the 11,521 captured pairs count as complete rewrites.",
  "answerSummary": "Models do not just tidy phrasing. They often replace most of the user wording with a more search-shaped query, changing the actual terms used for retrieval.",
  "keyFacts": [
    {
      "label": "Average similarity",
      "value": "25.24%",
      "detail": "Average Jaccard similarity between prompt and search query.",
      "claimId": "query-translation:similarity"
    },
    {
      "label": "Complete rewrites",
      "value": "31.85%",
      "detail": "3,670 pairs fell into the complete rewrite bucket.",
      "claimId": "query-translation:complete-rewrite"
    },
    {
      "label": "Prompt-query pairs",
      "value": "11,521",
      "detail": "Captured prompt-to-query transformations.",
      "claimId": "query-translation:pairs"
    }
  ],
  "evidenceTable": [
    {
      "label": "Average similarity",
      "value": "25.24%",
      "note": "Average Jaccard similarity between prompt and search query."
    },
    {
      "label": "Complete rewrites",
      "value": "31.85%",
      "note": "3,670 pairs fell into the complete rewrite bucket."
    },
    {
      "label": "Prompt-query pairs",
      "value": "11,521",
      "note": "Captured prompt-to-query transformations."
    }
  ],
  "whyItMatters": "Relying on raw user prompts for keyword research is insufficient. Teams must measure and optimize for the post-rewrite queries that models actually use to fetch context.",
  "whatToDo": [
    "Write for the query the model is likely to search, not just the wording a human might type first.",
    "Add current-year, format, and category framing where it genuinely helps the page answer the query.",
    "Make sure competitor and category language is present when the model is likely to inject it during retrieval."
  ],
  "faqs": [
    {
      "question": "How many prompt-query pairs were analyzed?",
      "answer": "The study captured 11,521 prompt-to-query transformations."
    },
    {
      "question": "What is the average similarity between a prompt and a query?",
      "answer": "The average Jaccard similarity between the prompt and the search query is 25.24%."
    },
    {
      "question": "How often do complete rewrites occur?",
      "answer": "Out of the analyzed pairs, 3,670 pairs, or 31.85%, fell into the complete rewrite bucket."
    }
  ]
}