Large Language Model Search Optimization Tool

The Vendor is required to provide a Large Language Model (LLM) search optimization tool or service that improves how institutional content is discovered, summarized, referenced, and represented within AI-generated search and answer systems. Also referred to as Generative Engine Optimization (GEO) or AEO (Answer Engine Optimization).
- The institution recognizes that traditional search engine optimization (SEO) is a foundational requirement for effective digital content discovery.
- Solution must therefore:
• Treat SEO as the baseline layer for discoverability and content health
• Build upon that foundation to support LLM-driven discovery and generative AI answers
• Provide measurable performance data
• Deliver actionable, data-driven content recommendations
• Support sustainable, non-technical workflows appropriate for a public institution
- Users increasingly discover information through:
• Traditional search engines
• AI-powered search interfaces
• Generative AI assistants and conversational tools
- must ensure that official, accurate, and current information is:
• Discoverable through standard search engines (SEO foundation)
• Correctly interpreted and summarized by LLMs
• Cited or referenced appropriately in AI-generated answers
• Continuously monitored and improved over time
• Securely managed and governed across traditional search engines and AI-driven answer systems
- Objectives
• Establish and maintain strong SEO fundamentals across institutional content
• Improve accuracy, consistency, and authority of AI Ǧgenerated answers
• Increase the likelihood that LLMs surface official institutional sources
• Identify gaps, risks, and outdated content affecting AI answers
• Measure content performance beyond page traffic alone
• Provide clear, prioritized, actionable recommendations content teams can implement
• Enable sustainable optimization with minimal technical effort
- Foundational SEO
• Analysis of search intent and high-value topics
• Identification of underperforming or unclear content
• Page-level content improvement recommendations
• Support for clear content structure, hierarchy, and authoritative signals as a baseline for optimization
• Monitoring of search visibility and engagement trends
- LLM Search Optimization
• Identification of topics commonly answered by AI systems
• Visibility into how institutional content is used and summarized by AI systems
• Detection of inaccurate, incomplete, or conflicting AI answers
• Alignment of content with conversational, question-based queries
• Reinforcement of official institutional messaging within AI responses
- Performance Measurement and Analytics
• Topic-level performance across traditional SEO and LLM-based discovery
• LLM visibility tracking showing how often institutional content appears in AI-generated answers
• Citation monitoring identifying when and how AI systems reference university content
• Analysis of sentiment or tone within AI-generated answers, including identification of negative, misleading, or incomplete narratives and linkage to underlying content gaps for corrective action
• Insights into when users receive AI-generated answers instead of clicking through to institutional web pages (zero-click behavior)
• Surfacing and prioritization of high-risk or high-impact content areas requiring attention
• SEO baseline metrics such as crawl errors, indexation status, and page-level performance
• Trend analysis showing changes in visibility, accuracy and overall performance over time
• Peer benchmarking comparing performance against similar institutions or competitors
• Role-appropriate dashboards designed for non-technical users
• Exportable reports suitable for leadership, governance and compliance review
- Content Gap, Risk, and Quality Analysis
• Topics lacking authoritative institutional content
• Content health scoring including readability, metadata completeness, and structural quality
• Content frequently misinterpreted by LLMs
• High-risk subject areas such as admissions, financial aid, and policies
• Redundant or contradictory content that confuses AI systems
• Outdated content contributing to inaccurate AI-generated answers
- Workflow, Usability, Governance and Responsible AI Use
• Transparency into why recommendations are made
• Human-review and approval workflows
• Content governance workflows supporting versioning, auditability, and departmental ownership
• Change management support to help staff adopt new AI-driven practices
• Responsible use of AI aligned with institutional values
• Avoidance of vendor-exclusive or proprietary content dependencies
• Non-technical user interface for managing workflows, recommendations, and configuration without coding