You probably know what you spent on paid search last quarter, but you probably can’t tell me how many of your enrolled adult learners first heard about your program from a coworker. That gap is structural. The marketing infrastructure most continuing education units have built was designed for a student who starts a Google search, fills out an inquiry form, enters a CRM sequence and eventually applies. That funnel is real, and it works reasonably well for traditional undergraduates and graduate students who spend a year or more researching programs. EAB found that more than 40% of master and doctoral candidates spent at least that long in their search.
Adult learners pursuing short-term workforce training give you about two months. They contact roughly two institutions. They decide quickly and with less information than any enrollment management funnel is built to serve.
And they find out about programs the way they find out about most things: from someone they already trust.
Discovery Happens Long Before the Click
Within the 2026 State of Continuing Education report, one finding tends to get noted and moved past: 30% of continuing education units lack dedicated marketing support, up from 14% the year before. The report frames it as a capacity problem, and it is, but there’s a second problem layered underneath it that capacity alone won’t fix.
Across thousands of graduate and adult learners, EAB’s research found that 40% of students first heard about their chosen program from a friend, colleague, family member or professor. A Gallup-Strada study spanning interviews with more than 22,000 U.S. adults found that half who attended college said their social network—friends, family and community contacts—was the single most common source of advice about their educational decisions, ranked above counselors and above institution websites.
For the workforce-training population specifically, it goes further. The Harvard Kennedy School Project on Workforce analyzed more than 75,000 programs listed on the WIOA Eligible Training Provider database and found that public program data is so sparse and inconsistent that enrollees effectively can’t make an informed comparison. They rely on career center counselors, on what friends and family have done and on referrals from providers themselves. The 2,400-plus American Job Centers across the country, the WIOA caseworkers inside them, workforce board liaisons—those relationships function as the actual discovery layer for a significant portion of the students continuing education units are now trying to reach.
None of that generates a click.
AI Is Making Invisible Influence Even Harder to Measure
The click was already a weak proxy for this population before AI search entered the conversation, and it’s getting weaker fast. Zero-click search, where someone types a query and gets an answer directly from the results page without visiting any site, stood at roughly 58% of U.S. searches in 2024. When it came to education-related queries, tracking February 2025 through February 2026, found AI Overviews appearing on 83% of them. up from 18% at the start of that period. A UPCEA and Search Influence study found that 50% use AI-powered search tools at least weekly, and 79% read Google’s AI-generated overviews when they appear. The prospect is forming a shortlist and a credibility judgment inside the AI response—before your website loads.
Tracking actual referral traffic, spanning January through August 2025 across thousands of Fortune 100 brand sites, found that AI search accounts for less than 1% of referrals and drives near-zero direct conversions. It functions as a top-of-funnel research channel whose influence is real and growing and systematically invisible to last-click attribution. A learner who heard about your program from a coworker, confirmed it via an AI Overview, then searched your institution by name will show up in your analytics as a branded search. The coworker gets no credit. The AI gets no credit. The referral network that actually did the work gets no credit.
When Measurement Drives the Wrong Decisions
Here’s where the measurement mismatch becomes costly. Recent data indicates that 93% of higher ed marketers track program web traffic, and 89% track traffic source. Fewer than 60% have visibility into how leads perform after they move from marketing into enrollment. Only 43% track cost per enrolled student.
The sector is heavily instrumented for the click and thinly instrumented for the thing the click is supposed to predict. When the click is degrading as a signal for your fastest-growing student population, the measurement problem becomes a resource allocation problem. Budget follows metrics. Clicks and inquiry volume as primary KPIs mean budget follows clicks and inquiry volume, and the channels that drove enrollment get counted as untracked organic traffic.
The average cost per enrolled student for online and continuing education programs runs around $2,849, according to Search Influence and UPCEA data. That number almost certainly undercounts what acquisition actually costs because it doesn’t capture the value of word-of-mouth referrals that never entered a paid funnel.
What Needs to Change
The fix is a reorientation, not a platform purchase. Measurement comes first. Before spending more on paid search, add a single question to your application: How did you first hear about us? Run it for two enrollment cycles. You’ll almost certainly find that friends, employers, workforce intermediaries and community organizations account for a larger share of enrolled students than they account for in your marketing budget. That gap is the reallocation opportunity.
The second shift is toward the channels the data consistently validates for adult enrollment: employer partnerships with genuine cobranding, formal relationships with NCWorks career centers and American Job Centers for workforce-aligned programs, and alumni and current student referral programs that give enrolled adults a structured way to tell people in their networks what they found. These are not supplemental channels. For the adult learner population, they are primary channels running without budget or measurement.
The third shift is structural visibility. BrightEdge found that 34% of AI citations pull from sources institutions can influence through earned media and public relations. Nearly 10% come from LinkedIn and Reddit. User-generated content sites are cited seven times more often than general business sites. Institutions whose reputations live primarily in paid ad copy and polished admissions pages are building for a distribution model that AI search and adult peer networks both discount. Institutions that earn citations in workforce board newsletters, regional business publications and community Facebook groups—not through campaigns but through genuine presence—will show up in AI responses and in the conversations that actually move adults to act.
A Problem We Share
I want to be honest about where this sits at Forsyth Technical Community College. We’ve built student-facing tools, including AI chatbots and digital intake experiences, that have meaningfully reduced friction for students navigating our programs. We track our digital channels reasonably well. But our measurement of the word-of-mouth and intermediary networks that likely drove a significant portion of our adult enrollment is thin. That’s a gap we’re working on, and I’m writing this because I suspect it’s a gap most of us share.
The Next Competitive Advantage
The adult learner population is not emerging. It’s here. It was found that 16.7 to 19.7% growth in nontraditional undergraduate students in fall 2024, even as traditional freshman enrollment dropped 5%. It’s estimated that roughly 42 million adults between the ages of 25 and 64 intend to enroll in education or training within the next two years. These are people who’ve already decided further education is relevant to their lives. The question is whether they find out your program exists before someone else’s does.
Most of them won’t find out because they clicked an ad at the right moment. They’ll find out because someone they trust told them. Whether that referral lands on your institution depends on whether you’ve built the presence and relationships that make your name the one people pass along.
That’s slower work than optimizing a Google campaign. It’s harder to report to a cabinet that wants to see cost-per-click trending down. But it’s the work that matches how the students you’re trying to reach actually make decisions. The signal exists. We just can’t see it yet.