AI at the Gates : Reshaping Enrollment and Financial Aid

Sathish:

“Welcome to AI Talks from BinaryWorks, where we discuss technology trends and innovation in higher education. We are joined by Aaron Basko, Chief Multiplier at Mack Insight Group, who has worked extensively in enrollment and financial aid. Aaron, thanks for joining.”

Aaron:

It’s my pleasure to be here. Thanks for having me.

Sathish:

“You have led enrollment and financial strategy across multiple institutions. What drew you to higher education and enrollment management?”

Aaron:

Ironically, I had a really good admissions counselor. Someone from my alma mater came to visit my school and recruited me. Years later, when I finished my degree and needed career advice, I went back to him. He encouraged me to try enrollment, so I got involved in admissions, and that is how my career started.

Sathish:

“What keeps you passionate about the industry after all these years?”

Aaron:

I believe education is a key to human flourishing. We want to better the lives of as many people as we can, and education is central to that. In my work at Mack Insight Group, we often say the best way to help people flourish through education is to make sure the institutions providing it are as healthy and strong as possible. That is what I focus on.

Sathish:

“How has AI changed the enrollment landscape and the way universities use it?”

Aaron:

It is an evolving process. We started with chatbots, and they have become more intelligent. Instead of just texting with students, they began reaching out, even calling students, which gives you amazing speed to lead. Now it is going further, into analysis of what students are doing and how they behave. One of our partners who is deep in AI describes it as moving to a thinking level, where AI becomes a strategic advisor. For a counselor, it advises how to connect with students better. For an enrollment leader, it pulls data in ways you never could before and helps you decide what is worth sharing with a group or your team. It helps you find the most relevant insights from week-to-week changes so your people can do what they need to do.

Sathish:

“Discovery is moving from Google to ChatGPT and Claude. Should universities optimize to get listed there, or will it happen organically?”

Aaron:

Institutions will need to change how they position themselves. It is now common for students to ask AI for a list of schools in the western United States with a certain major, ranked by cost. They are using that instead of traditional searching. So you have to think like a student and anticipate their questions if you want to appear on those lists. That is the passive side: being optimized to work with AI tools so you get in front of students in good ways. The active side is using AI to identify which students your outreach will actually make a difference for. It is a two-way street.

Sathish:

“Are enrollment teams already using these tools, and how will that change their work?”

Aaron:

It will change the role of the average admissions or enrollment professional. We will use AI to rewrite much of our communication flow and make it more responsive as students take certain steps or interact with us. It will also inform our conversations, helping us know whether now is the right moment to call a student and whether we have built enough of a relationship to make that call. Personal contact will be super important, but it will be personal contact informed by AI. Hopefully most interactions become more effective, with an advisor on your shoulder saying, the data tells us to do this, or this is how you need to reach out.

Sathish:

“Are enrollment teams using automation now? Are there tools they should try for an immediate productivity boost?”

Aaron:

Enrollment offices have been working on automation for years. Some enrollment pools include thousands of students, so how do you decide which ones to connect with deeply and which lightly? AI will help us take the next step. We have always struggled with speed to lead. It is hard to follow up in real time or near the time of inquiry. Sometimes we never get back to students, and when we do, it is often delayed. Students expect quick responses, and AI can provide them. I have seen tools where, as a student browses the institution’s website, AI tracks where they go and curates information: it looks like you are interested in this, would you like this as well? Then it asks if you would like a call right now, and the AI makes the call, giving the student that immediate response they want.

Sathish:

“Which parts of enrollment are benefiting most from AI automation: financial aid, student support, something else?”

Aaron:

All of those are probably in the process of being automated. Beyond customer service, AI will have a huge impact on modeling and prediction: deciding which students are most likely to enroll and what they need to enroll. It is already challenging our assumptions. For years our models told us which interaction was the highest predictor of a student’s decision. With more sophistication, we can see that maybe it is not the driver. Maybe it is just an outcome, not the cause. AI also helps us track trends week to week, which is difficult for enrollment offices, especially in the critical last six weeks before students decide. It speeds up analysis so you can make quick adjustments. And it will help significantly in financial aid, with efficiency in packaging, predicting what students need, and helping them self-serve in ways we never could.

Sathish:

“Are students choosing online courses over the traditional campus experience? Should colleges move more toward online education?”

Aaron:

The trend is in that direction, though not immediately. There is still a strong base of students who want to live on or near campus and interact in person. I think we have struggled to make the online experience equal to the in-person one, and there is a lot of data showing we are not delivering it as well as we could. Graduate education has moved online quickly; it is the preferred method in much of that space. Undergraduate is moving more slowly but in the same direction. We often have a quality gap, so tools that improve quality and the student experience will make that possible. But almost every institution needs an online option in its repertoire to reach the growing market of students who only want to learn that way.

Sathish:

“Could financial aid be automated, perhaps with a 24/7 AI financial aid counselor, or should it stay in person?”

Aaron:

I see many use cases for AI in financial aid. It is sensitive, heavily regulated, and institutions are accountable to the federal government for doing it right, so the area adopts AI more cautiously. Financial aid has struggled with automated packaging for years. The tools exist in most systems I have seen, but they are clunky. They need huge upfront effort from financial aid professionals to set up, and even then it is not fully automated. People still download files and push buttons. AI will need training and oversight, but it could greatly speed up automation and run quality control checks on packaging.

The bigger opportunity is helping students self-serve and navigate the journey. So many students get lost in financial aid, and counselors do not have time to call everyone to ask why a student is stuck. AI can give students an advocate. I picture it walking in front of you, opening the door, and helping you talk to the right person more intelligently, acting as a liaison for the student and family. It can keep asking, why are you stuck here, and either solve the problem, flag it for a real person, or tell the family it is time to talk to someone. That can help students get unstuck and move smoothly through the funnel.

Sathish:

“Is any tool doing this well today, or are you looking for one?”

Aaron:

We are actively looking at this at Mack Insight Group, since financial aid is much of what we do. With AI’s increased predictability, we are looking for ways to start the process much earlier for families, even before admission. Today institutions rarely engage families before they are admitted, which is far down the funnel. If you can predict what a student is likely to need, you can engage families earlier, pre-package aid, show them what it will look like if they are admitted, and guide them through the timeline. Some institutions are likely experimenting with this, but it is not industry-wide yet. It is still early. There have been critical shortages of trained financial aid staff who can manage the regulations, so tools that help them handle the volume will be an incredible benefit.

Sathish:

“Do regulations act as a bottleneck for financial aid AI tools?”

Aaron:

There are concerns. Compliance is much higher in financial aid than in admissions or other parts of enrollment, so you have to be more careful. But part of it is a dynamic difference. Admissions offices are always pushing for the newest thing to gain a competitive edge over other admissions offices. Financial aid offices do not have that competitive dynamic; they are more of a service area. They tend not to be early adopters and move more cautiously. As they see what tools can do, they will become interested. Companies serving financial aid will also have to confront this. And the more an institution sees financial aid as a driver of enrollment, the more likely it is to adopt new tools. Some institutions still separate recruitment and financial aid, treating aid as a packaging service. Institutions where the two work together are more likely to be early adopters.

Sathish:

“Retention also carries financial risk, especially for students on aid. Are predictive systems in place to identify students likely to leave early?”

Aaron:

This area has actually been a bit ahead. Some tools, which students may opt into, constantly monitor behavior or act as a chatbot companion that checks in: how is this going? Based on students’ feedback, they make suggestions or notify staff. These have existed for a while and will keep improving. It is a huge need, because as an industry we generally underperform on retention. We still lose many students to the bumps and bruises along the way.

I have worked a lot in retention, and you need to get students truly connected to the university. My hope is that AI breaks down barriers by facilitating an interaction or telling a student which office to connect with. But students often hit other, less friendly barriers once they get there. What I really hope is that AI drives students toward the right interactions with people on campus, because building that relationship is critical. We will see whether students can build a sticking relationship with the AI itself. In the short term, AI can connect students with the right people, knock down common pitfalls, make suggestions, and notify us who may have retention issues. It can even be honest with students: students with your current experience do not typically persist, so here are steps that make it more likely you will.

Sathish:

“Are there specific retention tools you use or are considering?”

Aaron:

I have worked with a few companies in this space, including EdSights, and I have thought about whether more could be built. I have spent a lot of time getting departments to work together, especially academic advising and financial aid, where there is usually a big barrier in how they collaborate on student success. There is a lot of room for new tools. I have seen a few models, but I think it could be much better.

Sathish:

“What percentage of students leave before completing their degree?”

Aaron:

Nationally, only a little over 50 percent of students earn an undergraduate degree in four years. I do not think that is very good. Some institutions do great work, but if an industry only serves about half its customers well, that is not great. It is also a huge opportunity. With fewer students and more competition for them, keeping more of the students you already have, instead of recruiting 40 percent more, is a huge ROI for institutions.

Sathish:

“Have the tools you have seen reduced that dropout rate? Are there early signs of improvement?”

Aaron:

There are examples of moving the needle, usually by two or three percent rather than ten. The biggest challenge is that so many factors are involved: is it the tool, the increased emphasis, or the institution improving its processes? The ROI can be hard to measure. If it works and you see the improvement you want, using all of them may help. There are many factors in a student’s decision to leave, and we have not done as much modeling on that as we could. Some companies have spent about a decade analyzing ten years of student data to find indicators of persistence and the obstacles in students’ way. As we move forward, tools will get more sophisticated, do some outreach for us, and challenge our assumptions with data.

Sathish:

“Will universities develop personalized tools that guide each student through classes and tasks to improve engagement?”

Aaron:

I think AI will become a companion, on both the administrative and student sides. Our partner described an enrollment manager asking the AI: I have a board meeting in three hours; based on our data, what should I share? That is an amazing tool, pulling the latest data and advising how to present it to the right audience.

It will do the same for students, creating a more curated journey. Instead of asking friends, which is usually a bad source of information, or waiting for time with an advisor, they will ask their AI companion how to register on time or pay their bill. That creates some dependency, but it is just the next iteration of asking Google or Siri. My last institution was experimenting with a chatbot that went beyond the public website. Students could log in and ask it to analyze their own record: what classes should I take, or what do you recommend based on my financial data? It is already becoming a counselor and source of information.

Sathish:

“Could these tools help smaller colleges compete with large universities by offering a similar experience?”

Aaron:

If done right, it could be an equalizer. The playing field is more unbalanced now than at any point in my career, with big institutions holding many advantages and smaller ones many disadvantages. If smaller institutions are nimble, they could use this technology to shift things back: offering a highly personalized experience, stronger service, and more availability. When institutions ask me what to think about AI, I say it is a great opportunity, but you need to build it specifically for your institution. It is not one-size-fits-all. If smaller institutions embrace it, they can find the students they need, gain visibility, and provide a higher-level experience that also helps retention.

Sathish:

“You have deep experience in international recruitment. Is AI changing international student recruitment?”

Aaron:

International is a passion of mine, and I think we are missing a huge opportunity. Institutions are nervous after dips in international enrollment and concerns about the regulatory environment, so they have backed away. That is the wrong approach. You should question whether your methods still fit the new environment. For example, people worry about students getting visas. I am working on a project with a couple of institutions offering online dual enrollment to international high school students. Few are doing that, and it is a market of thousands and thousands of students. Institutions that think differently can fill seats and work around current circumstances.

AI helps you understand those markets, choose them intelligently, and speak to them better by translating your message into the local vernacular. With the domestic market so tight and students applying to the same schools, institutions need new sources of students, and those students are out there. They just will not come the way they used to. There are also opportunities with international organizations. A connection of mine is networked with hospitals in one country. They do not want to send employees away for a four-year bachelor’s degree, but a one-year certificate to upskill them is a wonderful opportunity. It just requires thinking differently.

Sathish:

“Are universities promoting directly in international markets, or connecting through organizations like those hospitals?”

Aaron:

It is a mix. Traditional promotion is hard, and AI could help fine-tune messaging for the right markets and micro-markets, so you are not advertising to everyone, and predict where promotion matters based on where you have historically enrolled students. But much of it is connections: finding someone who can bridge the two sides. That is a lot of my international work. Institutions tell me what they want, and I help them find partners and sources of students. There is a need on both sides, so the students, the organization abroad, and the institution here all win.

Sathish:

“Are universities missing opportunities in international recruitment?”

Aaron:

Yes. When people get nervous, they tend to throw the whole thing out instead of adjusting tactics. Many institutions are backing away from international students because it seems like more work or the old methods no longer work. But that market is enormous, probably ten times the size of the US market. It is about positioning yourself correctly and delivering service that makes sense in those markets. Too many institutions became dependent on one or two countries, and when things changed there, they concluded international does not work. Instead, shift focus to other countries. I have been doing great work in Latin America, a market that is doing very well, and most institutions are not taking advantage of it.

Sathish:

“Many university leaders follow this podcast. Where should they invest in AI first?”

Aaron:

It is hard as a leader. You feel you should know everything, and people ask what you are doing with AI when you may not even know what is out there. For many leaders, the first step is becoming aware of what exists, and there is no shame in saying, I need to learn. For many institutions, the best move is finding a trusted partner in this space. You do not want to buy a random solution without knowing if it fits. You want to build something specific to your institution and the messages you want to send. There are obvious steps, like a great chatbot or a more advanced tool that reaches out to students for speed to lead. But predictive modeling needs one kind of tool, and making your financial aid office more efficient needs another.

It is also an opportunity to ask how AI can help find the leaks in the boat. Where are we dropping the ball? Do not be afraid of that; everyone drops the ball somewhere, and nobody has a perfect process. First, use AI to identify the leaks. Second, use it to close the gaps. If you have plenty of prospects but cannot follow up fast enough, AI can help. If you do not know whom to contact or what conversation to have, AI can help. If you lack visibility into what worked in the past and which markets or strategies deliver results, AI can help. A lot of it is asking questions, and knowing whom to go to for answers is a big deal.

Sathish:

“AI can hallucinate, and students make decisions based on its answers. How should universities approach AI governance?”

Aaron:

It is easy to forget the problem exists. It works most of the time, so you assume it always works. You have to keep reminding people it is not foolproof and has failures and limitations, and always put human checks and balances in place. In financial aid, it is critical to have human checks. AI can do the heavy lifting, but someone must quality control it, because precision matters so much. In other areas, precision is less important. If AI suggests which students to call today and gives you a couple of wrong names, you made a few extra calls. No big deal.

So with hallucinations, ask how big the downside is. If it is like the calling example, it does not matter if it is occasionally wrong. If it occasionally reports something incorrectly to the federal government, that is a problem. You are always weighing risk and reward. Where risk is low, let it run. Where risk is larger, assume it may be wrong some percentage of the time and watch for those errors. Do that risk-reward analysis everywhere you use it.

Sathish:

“There is a fear AI will affect jobs. Will it create fear in enrollment organizations?”

Aaron:

I think it will shift jobs rather than eliminate them in number, though maybe it will. Mostly it will change what we do. I will show my age and compare it to when email first appeared. We thought it would take so much work away and give us extra time. Instead, we have lots and lots of email. Transitions like that create shifts. Maybe fewer people needed secretaries to take dictation, and mailing budgets fell, but new jobs appeared because the technology let people do new things. Career services, which is near to my heart, is a good example. You will need fewer people for resume reviews, because AI can write a resume nearly as well. But you will need more people teaching soft skills: how to network and think strategically about long-term career moves. AI will increase demand for higher-level, intensely relational skills. The human interactions you do have will become much more important and need to be much higher quality, while AI replaces some lower-level interactions.

Sathish:

“How will enrollment and universities look in three to five years?”

Aaron:

We will see some pretty dramatic changes. In many ways, this part of higher education has been waiting for something to help it get unstuck. We have been stuck in old models of finances and recruiting, doing the same things for years. AI will challenge institutions, but it also gives them the tools to adapt. In three years, it will change how we do the work: handling application volume and outreach differently, and personalizing in ways we could not before. Around five years, it may change how people do business. The application process might change, because so much of what students submit can be produced by AI. Do you still need to collect it? AI may also advise institutions to change their model: you do not belong in this space anymore; strategically, move into this one. In three years it helps us do what we do better. In five years, it could change what we do. It could be transformational.

Sathish:

“What is the biggest pain point or opportunity in enrollment and financial aid that could be completely disrupted?”

Aaron:

There are several. One is that students are fleeing to the most selective, big-brand institutions because of concern about whether degrees lead to jobs. The data does not support that. The quality of a student’s experience at an institution is far more predictive of job success than the institution’s selectivity. But that is a huge problem, because three quarters or more of US institutions are not highly selective, and they are struggling to rethink their model. Since roughly the 1960s or 1970s, the model has been to recruit more students every year and raise tuition slightly. What happens when there are fewer students and they are less willing to pay tuition increases? That model breaks for three quarters of our institutions. You see budget cuts and tightening, and institutions unwilling to invest in their own future precisely when they need to most.

If technology can help us get outside that model and think about revenue differently, that is the opportunity. In every meeting at Mack Insight Group, we say we want to be the ROI company for institutions. We are not satisfied with a two- to three-times return; we want our clients to see ten times. We worked with an institution in California and helped them recover $5.6 million in pre-collections over two years. Before sending students to collections for unpaid bills, we reached out, gave them options, and they paid part of their bill. People think of that money as a throwaway. AI will help find those opportunities. It will say: your boat is leaking, here is where. You just lost six million dollars because you did not help people pay their bills. Let me help you target the people most likely to pay. Hopefully it will also suggest going outside your normal market, like international dual enrollment, where there is a huge market and few competitors. We have been stuck thinking it is about more students and the revenue will take care of itself. When more students are not available, you have to shift your mindset. This is about revenue, which lets us deliver on our mission for students. The more AI points out these obvious problems, the more it will help us transform.

Sathish:

“Any advice for future enrollment leaders?”

Aaron:

See it as an opportunity. It is easy to see AI as a threat: it will hurt jobs or make the process less personal. Those downsides are easy to see because we know our work. What is harder to see is what we do not yet know will happen, and there is opportunity there. Use AI as your companion and ask, where am I missing? Keep using your gut instinct, your training, and your best people, and reinvest in relationships. But use AI to scan your environment and find where the issue is not about trying harder, but about trying differently. AI can become the bridge between us and our students, making those interactions much more fluid and effective.

Sathish:

“Thanks for your time, Aaron. With so much changing, I hope we can reconnect in six months, when new tools will have changed things even more.”

Aaron:

Absolutely, I think that is going to happen. Thank you for having me.