The first real conversation a prospective graduate student has about a program now often happens without the program in the room. Before a learner submits an inquiry form or speaks to an admissions counselor, many have already asked an AI assistant what a degree is worth, weighed it against alternatives, and reached a tentative verdict. By the time they raise their hand, the decision is largely made. They want confirmation that they have chosen well, and they have little patience for a sales pitch.
That shift sits at the center of “Working Theory,” a new newsletter from Kees Bol, chief executive of the education company 2U. Bol describes the project as a running set of observations from “the intersection of learning, work, technology, and leadership,” and titled his opening issue “When customers arrive informed.” The questions he says he keeps hearing from leaders are familiar to anyone running an institution today: how to adapt without losing what makes you valuable, how to make good decisions when the future isn’t clear, and which changes deserve real attention rather than reflex.
How AI is changing the online education funnel
For most of the past decade, colleges and their enrollment partners treated the early stages of the student journey as a teaching problem. A prospect didn’t know the difference between a master’s in analytics and one in data science, so the institution explained it. Large language models have quietly removed that step. A learner can now get a competent, side-by-side comparison in seconds, which means the first impression is increasingly formed before the school gets a word in.
What a prospect wants from the institution changes as a result. The value of being walked through the basics drops toward zero when a chatbot does it instantly and for free. What rises in value is help making a confident choice: clear evidence about outcomes, honest answers about fit, and a credible signal that the program delivers what it promises. The learner arrives further down the path and is less patient with anything that sounds like a pitch.
Why discoverability now matters as much as recruitment
The practical consequence is that visibility does work recruitment used to do. If a learner builds a shortlist through AI-mediated search, a program earns its place on that list well before any human conversation, or it never enters the running at all. Accuracy, credibility, and presence in the sources these systems draw from become enrollment infrastructure rather than marketing polish.
This is the uncomfortable part for institutions that measure success by traffic to their own websites. The decisive moments increasingly happen somewhere else, inside tools the institution doesn’t own and can’t fully see. Influencing those moments calls for a different muscle than running a campaign: being consistently findable, trustworthy, and useful across the channels and content that shape what AI surfaces.
The scale behind 2U
2U operates edX, the global learning platform, and works with more than 250 university and company partners on over 5,300 programs reaching a learner network above 100 million. In Bol’s account, that footprint matters less as a headline number than as a source of pattern recognition. Across degree and non-degree programs, the company can see which formats hold attention, which markets are growing, and which credentials lead to jobs.
Bol made the point plainly in an interview shortly after taking over in early 2025. The company’s size, he argued, gives it “the ability to look across the market and portfolio and see trends, opportunities and market signals” that a single institution can’t easily assemble alone. In a market where learner behavior is changing faster than any one campus can track, that vantage point is the asset.
From vendor to force multiplier: 2U’s partnership thesis
When Bol spent his first months meeting university leaders, the message he reported hearing most often was a warning. As he recounted, one leader told him: “We don’t need another vendor. We need a strategic partner who understands where education is heading and can help us get there.”
His answer is a deliberate reframing of what an online program partner is for. “Our most successful partnerships are with universities that see us as a force multiplier, not a replacement,” Bol said, describing partners the company can help “move faster, reach further and create more impact than either of us could alone.” The logic tracks with the informed-learner shift. As institutions build their own digital capabilities, the scarce resource is no longer execution. It is foresight: knowing which way learner demand is bending and showing up there first.
What arriving informed demands of universities
Bol’s through-line is that technology should sharpen human judgment, not stand in for it. He has pointed to a line from a university provost that he says captures the idea: “The magic happens when technology enhances human connection rather than diminishing it.” Applied to the informed learner, that means using AI and data to identify which prospects are a genuine fit and to support them earlier, while keeping a person in the loop for the decision that actually matters to someone’s life.
The institutions that adjust to this will look different on the surface and familiar underneath. They will spend less effort explaining themselves and more being worth finding. They will treat credibility as something earned in public, across the platforms where decisions now form, rather than asserted on a brochure. And they will lean on partners not to outsource the work but to see around corners. Bol’s wager, and the premise of his newsletter, is that the learner who’s already done the homework is the clearest argument yet for getting that model right.
