Finding the Right Balance with AI Personalization in Admissions

Admissions counselor Sam sends an email to prospective student Sarah that references a volunteer service award she received two years ago and notes how she might enjoy similar volunteer opportunities at the university. Sarah was recognized for her volunteer service in her hometown paper.

Impressive?

Or Unsettling?

Today’s AI tools can gather much more public information from multiple sources than any admissions counselor ever could. This creates opportunities for hyper personalized information to be used in communication plans and throughout the recruitment cycle. However, it also poses several important questions. What is acceptable information to add into a student’s record? Where is the line about what is too personalized?

As AI becomes more sophisticated, admissions counselors find themselves at the center of an important ethical conversation. How far should universities go in implementing this technology when messaging prospective students?

Benefits Of Personalization

Personalization has value in the communications cycle, if not pushed too far. Communications and admissions professionals regularly hear from students about too much generic outreach that doesn’t relate to a student’s interest. Personalization is one way to address this, which ultimately improves the experience of prospective students.

Additionally, AI personalization, when done well, can help enhance communication by adding in thoughtful touches that resonate with students. This kind of content has shown to break through the clutter of messages prospective students receive and helps improve email and other communication response rates.

At the root, using personalization is about striking the right balance of including personalization while not pushing the technology so far that it becomes creepy or uncomfortable for prospective students.

When Personalization Crosses the Line

Not all data that an admissions counselor can access through AI tools was intended to be used in college decisions and college communications. AI tools scrape data from local papers, high school websites, social media posts, and more. This means there is lots of public data about prospective students, but not all of it has been shared with colleges and universities.

When trying to decide which of this publicly available data should be used in communicating with prospective students, a good place to start is thinking about the appropriateness of the content from the student’s point of view. If prospective students would be surprised or wary with that information included in a message to them, it is probably best to leave that information out.

This requires a mindset shift for admissions counselors. Instead of focusing on what data is publicly available, it is crucial to think about whether it is ethical to use that data.

‍

Example: “I noticed that you mentioned an interest in chemistry during your campus visit. We have a new research lab coming online in this area, and I thought you might like knowing about it.”

In this example, the student shared something during a visit and that information is referenced in a follow up communication when a related update is announced. This seems reasonable and would likely delight the student to receive a message that relates to their interest area they shared with the university.

‍

Here is a scenario that is a bit more problematic.

Example: “I noticed that you post a lot on your TikTok about science things. We have a new research lab coming online in this area, and I thought you might like knowing about it.”

While the video may have been public, the prospective student likely didn’t expect it to come up in a conversation from an admissions counselor during outreach. The student may feel as though the admissions counselor is digging a little too deep into social media and other locations to learn about the student.

‍

‍

Keeping the student perspective in mind, below are a few other scenarios to help guide decisions about whether data should be included.

Fair Game

• Information provided in an inquiry form
• Major or academic interests listed on the application
• GPA, test scores, submitted coursework
• Club, athletic, or academic competitions notes on an academic application
• Public achievements listed in resumes, essays, or scholarship applications
• Campus visit history or attendance at university admissions events
• Questions asked to an admissions counselor
• Interests the student expressed during a campus visit

May Cross the Line

• Content scraped from personal social media accounts
• Political views or activism not related to admissions
• Data inferred from browsing behavior on other websites
• Referencing family financial situations gathered from third-party sources
• Mental health, medical or disability-related content scraped from social media
• Old social media posts or photos that are dated
• Using AI to infer personality traits
• Combining multiple public sources to create a more robust student profile

‍

Consider the Trust at Stake

Another reason to be judicious when using hyper personalization is considering how students will perceive the message. If a student receives an unexpected message from an admissions counselor, while technically accurate, it could be unsettling and come across as “creepy”.

Receiving these kinds of message has the potential to impact the student’s willingness to trust the admissions counselors and others associated with the institution. That kind of risk goes far deeper than the communications cycle and the admissions process. It could have long-term impacts on prospective students’ enrollment decisions. Losing trust that the university has a students’ best interests top of mind is a serious outcome of overusing personalization. University admissions counselors and other leaders should take this risk seriously when considering how much AI to use in prospective student communications.

Don’t Diminish Human Responsibility

AI is a great thought partner in working with prospective students. It can help craft emails, suggest the best time to reach out, suggest topics that may resonate with students, and provide detailed data about students.

However, the human members of the team have a responsibility to ensure the content is appropriate, respectful, and aligned with institutional values. That work can’t be done by the AI technology, which is why it’s important to have human team members involved in decisions about how the communications are built and what information is okay to use.

At its core, the goal of personalization is to build trust. By using personalization in messaging, universities can show they know students, they are listening to their needs, and they care about ensuring students have success.

Today’s prospective students are aware that data is collected and repurposed in other ways. Institutions that are respectful of personal information and incorporate data in meaningful and authentic ways, build far more trust than those who overuse data.

A message that feels genuine and relevant leaves a strong lasting impression. While a message that feels overly intrusive may cause students to question the communication and the institution sending it out. When every interaction contributes to the institution’s overall brand perception, the risk of over personalization can be a significant one.

Closing Thoughts

Access to AI tools makes meaningful personalization possible. However, it also interjects an ethical risk that over personalization can damage institutional trust if admissions counselors send too many intrusive messages. The element of human judgment is critical in deciding whether personalization is appropriate in each se case.

So, let’s revisit Sam and Sarah. Sam referenced the volunteer award and connected it to volunteer opportunities.

Was that thoughtful outreach that aligned with Sarah’s passions?

Or did it leave Sarah questioning if the institution is sleuthing on her?

The answer depends on Sarah’s perspective. And that’s somewhat the point of the example.

AI can certainly personalize messages. But it is up to the student to decide whether that personalization strengthen trust and help a prospective student feel understood or is it an unsettling reminder of how much the institution knows about them. That answer lies in knowing the kinds of students at the university and what matters to them, something very human.

The institutions that will navigate this tension the best likely won’t have the most sophisticated AI tech stack, but they will be ones who know when to let technology assist and when to override the technology in favor of human judgment.