Online marketing is facing two challenges at the same time:
- People are discovering information differently
- We have less data on how they actually behave online
It is clear that people, and prospective students among the first of them, are using LLMs to inform their decisions. As marketers, we need to adjust and meet them where they are.
But as more interaction happens away from our websites and social media channels, we lose visibility. We have less first party data, such as web analytics, to base our decisions on.
What happens inside the chatbot stays inside the chatbot. Google still gives us keyword and impression data in Search Console, there is no real equivalent for Claude, ChatGPT, Gemini or Perplexity.
So if you want to understand how prospective students use LLMs in their decision-making, you have to ask them. Better still, observe them while they do their research.
And that’s exactly what OHO, a US agency and UCAS have done.
OHO ran a survey with a diverse panel of US-based respondents and carried out ethnographic research, observing prospective students as they researched colleges. Here is the full webinar recording.
UCAS surveyed 4,485 prospective students and ran 3 focus groups. The full report is not out yet, but some results were shared in this Wonkhe article.
The OHO video (0:56) is especially worth watching.
If you don’t have the time for that, here is what stood out to me:
AI as the first step
UCAS reports that 48% of respondents used AI to explore their options.
OHO asked respondents to rank different sources in the order in they used them during college research.
In that study, AI tools came first, closely followed by search engines. University websites and social media came after.
In other words, for many prospective students, the LLM is becoming the first step before they dive deeper into websites and social content.
Typical questions people ask AI
Prospective students use AI as a research tool. They provide it background information about themselves and what they are looking for, then use it to explore their options.
Common topics include:
- Comparing universities
- Exploring subject choices
- Entry requirements and admissions fit
- Career pathways and outcomes
- Location and lifestyle fit
- Costs and value
- Planning next steps in the application journey
- Summarising and simplifying information
AI vs. other information sources
AI platforms are becoming a common entry point and are used heavily in the research stage. When they are used, they shape the process and may influence whether your institution makes it into the consideration set.
But LLMs are not the whole story.
For final decisions, other sources still matter a great deal, especially the university website and what students find or hear elsewhere, including on social media.
The graph below answers the question:
Which of the following influenced you as you researched your college options?
It only includes the subset of respondents who had used AI.
Implications
The good news is that much of what helps with AI visibility also helps with the other major sources of influence, especially your website and social media presence.
Your website
Many of the questions prospective students ask AI are about topics that should already be clearly addressed on your website, such as:
- Admission requirements
- Available courses
- Costs Â
- Career pathways
If those answers are explicit, clear and accessible to AI crawlers, you are far more likely to be represented accurately.
Student ambassadors
Beyond your website, prospective students verify what they find through social media and third-party platforms.
LLMs often pull in outside sources for student perspectives and questions about student life. That means an engaged group of student ambassadors can help in more than one place, especially if they are active on platforms such as Reddit.
See you in 2 weeks!
This article first appeared as a Field Note
In the Field Notes, I share one practical idea you can try on your site or in your campaigns. Every two weeks. Get it in your inbox by signing up here.
