1.
The AI Revolution In Student Discovery
LLMs (Large Language Models) like ChatGPT and Gemini are rapidly becoming a primary resource for prospective students seeking information about colleges and universities.
ChatGPT alone now processes over 1 billion daily queries, a fivefold increase in just 12 months (as of June 2025, when this is being written).
Net Natives have long been pioneers in the use of AI, we were shortlisted for our first award Innovation AI in 2020 for our Akero data platform back in 2020 - 2 years before the launch of ChatGPT! Our vision is that we create opportunity by connecting our partners with every potential student and so it’s our job to ensure we are here to support you through this evolution in student marketing.
ChatGPT alone now processes over 1 billion daily queries
Understanding Share of Model (SOM)
SOM is how often your university appears in AI-generated answers, in other words, how much “space” your brand occupies in the mind of an AI model when users ask relevant questions in the form of prompts.
The reality is stark and binary: Colleges and Universities either appear in AI responses or they don't. There's no middle ground, there are no page 2 results.
Not all LLMs approach training their models in the same way. Some, like ChatGPT, are trained on static data up to a cutoff point (e.g. December 2023), so their answers reflect past learning. Others, like Google Gemini, pull from the live web, meaning your brand must be visible in up-to-date, authoritative sources to be included in responses.
AI Visibility & Reinforcement Learning
LLMs don’t just pull links, they generate answers based on patterns learned from huge datasets. And critically, these models are fine-tuned using Reinforcement Learning from Human Feedback (RLHF). This means the AI “remembers” which institutions, programs, and stories users find most helpful and trustworthy.
The more your university is referenced and positively engaged with, through queries, user feedback, and content quality, the more your brand becomes embedded in the AI’s knowledge base. This creates a compounding effect: better visibility leads to more recognition, which leads to even higher visibility in future AI updates.
Optimizing for SOM is about becoming the university name AI tools recommend, helping you win the attention of students at the very start of their decision-making journey.
How to Boost Your University’s Share of Model
Traditional SEO tactics of E–E-A-T* which impact organic search results are even more critical for AI Discoverability. But AI Discoverability requires more than just *Experience-Expertise-Authoritative and Trustworthiness techniques.
1. Optimize Your Website Structure for AI Crawlers
LLMs rely on well-structured, well indexed and crawlable content to find and reference information effectively and use in featured snippets or AI summaries. LLMs prioritise content from trusted sources.
- Your domain (.edu/.ac.uk) is the ultimate in “trustworthy source”. Ensure you have brand consistent, values-driven messaging across your website.
- Use clear, question-based headings (H2, H3) that match student queries
- Implement structured data (schema markup) like FAQPage, HowTo, and Review to provide explicit context
- Ensure your site is crawlable by AI bots with a clean robots.txt and XML sitemap
- Optimize meta titles and descriptions for inclusion in AI answer snippets
- Write short, scannable paragraphs with bullet points and lists for easy AI parsing
- Build logical internal links between courses, research projects, and alumni profiles to strengthen topical authority
- Improve site speed, mobile-friendliness, and security (HTTPS) for better user and AI experience
2. Create AI-Friendly Content That Answers Student Questions
AI models prioritize content that directly and clearly answers user prompts.
- Use natural language that matches how students ask questions (“What scholarships does University X offer?” or “Which College is best for Y”)
- Provide detailed FAQs, guides, and explainers that cover both big questions and related subtopics related to course content
- Include real data, like graduate employment rates or scholarship amounts, to build trust and answer specific questions
- Example: Instead of saying “Our careers service helps students get jobs,” say “95% of Computer Science graduates from University X found jobs within six months, with an average starting salary of $X.”
- Feature authentic voices: quotes from professors, alumni testimonials, and student stories
- Demonstrate expertise and authority by citing credible research sources
- Keep your tone human and engaging, AI favors content that’s genuinely useful and readable
3. Build Your Brand’s Authority Across the Web and Social Media
AI models learn from the entire digital ecosystem but places greater authority on more “trustworthy sites”:
- Keep your profiles up-to-date and consistent to your brand on ranking sites, Wikipedia, and academic directories
- Engage in discussions on platforms like Reddit and LinkedIn to share insights and build social proof (especially important to increase relevance in LLMs that use this “live data”).
- Publish thought leadership content and celebrate research breakthroughs and student success stories across trusted news and academic sites
4. Increase AI Discovery & Influence Reinforced Learning Opportunities Through Effective Advertising
While AI models don’t directly “see” paid ads, effective, cohesive and consistent brand messaging reinforces your institution’s value based messaging to inspire the student prompts.
- Understand your correct brand messaging and stay consistent wherever a student will become aware and engage. Know what you stand for and join that up with creative messaging that will resonate and inspire.
- Focus on creative excellence. Clear, emotional, and value consistent creative not only perform up to 4x better (according to Kantar Research) but also reinforce a seamless messaging. Leverage the full extent of innovative engagement opportunities on the platforms, from augmented reality to dynamic rich media.
- Engage throughout the journey. Think brand safe, authoritative display on trusted media sources combined with effective TV and integrated with Out of Home that tells a consistent brand message that will inspire prompts with your brand included.
- Integrate authentic influencer content into media strategies across social and content channels to add credibility and human connection
- Create content campaigns that impact reputation by promoting research breakthroughs and community impact. This can influence opinion which enriches the informational ecosystem that AI systems learn from, namely Institutional Ranking positions.
- Be contextual. Adverts resonate when they appear in the right context. Insert yourself into the story when students are engaging with content that aligns with their interests; whether they’re reading about a related subject, listening to music in a fitting genre, or watching TV programmes with thematic connection.
- Google search still dominates discovery and is essential for converting high-intent audiences. Run search ads targeting relevant program keywords alongside value based messaging against competitors
5. Collect, Connect & Visualise Your Data To Measure Impact
Connecting your data to know if your SOM efforts are working and what ROI your marketing energy and advertising spend is generating.
- Track referral traffic and branded search volume from AI tools and traditional search engines
- Use consistent UTM parameters within your advertising and your analytics and integrate with your CRM
- Build a "First Touch to Enrollment" View to map inquiries and applications to measure and improve outcomes
5. Collect & Create Your Own Data Assets
Share of Model visibility is impossible to measure directly without collecting your own data assets (inquiry data from students and how they progress through the journey). This limitation is compounded in our cookie-depreciated privacy age.
Establish your "Value Exchange". Why should a prospective student or fundraiser give you their data? How can you improve the journey from data exchange to a converted student or fundraiser?
6. Measure Human Sentiment
AI will process massive datasets and detect patterns to generate content and predictive modelling. But we need to validate the findings with lived experiences, emotional nuance, and contextual understanding that can only come from real people.
- Collect human feedback through surveys on brand recall and perception
- Track prompted and unprompted brand recall over time to validate data with lived experiences.
