Why Your Enrollment Team Is Buying the Wrong Thing (And It's Costing You the Class)

Here’s the deal.

Ask your enrollment team what their cost per lead is. They’ll tell you instantly.

Ask them what their cost per start is. Watch the room go quiet.

That’s the tell.

A purchased inquiry looks like work. It hits the CRM. A counselor gets a task. The dashboard moves.

Then reality shows up.

Half those names never pick up. The ones who do? They’re sitting in three other schools’ inboxes. And your class still comes in light.

The product you bought was a form fill.

The product you needed was belief.

What a “Lead” Actually Is in 2026

Pay-per-lead still sells well in higher ed. Why? Because the unit is simple. You pay. You get a name.

For most programs, that name costs $140 to $200. Specialized grad, healthcare, law, and executive programs? Try $1,000+ per contact.

Now here’s what you didn’t buy:

Exclusive ownership. Shared leads go to three to five schools. At once.
A student who asked for you. They asked a matching site. A ranking page. A “see if you qualify for aid” form.
A warm conversation. About half of purchased leads never respond after delivery.
A conversion rate that matches your own traffic. Purchased leads convert around 2%. Your own first-party inquiries? Closer to 4–5%.

And here’s the kicker: PPL now accounts for roughly 4% of total enrollments — down from more than 8% in earlier cycles.

The channel didn’t get smarter. The students did.

They research on your site. On YouTube. On Instagram. On TikTok. And now — inside AI answers.

The aggregator form is what they fill out when they’re shopping. Not when they’ve chosen.

The Math That Hides in Plain Sight

Cost per inquiry looks tidy. Cost per enrolled student does not.

Recent benchmarks:

Undergraduate inquiries: ~$128. Enrolled students: ~$1,505.
Graduate inquiries: ~$157. Enrolled graduate students: ~$3,800.

If your team can’t split cost per start by channel, the cheap lead is hiding the expensive class.

Write that one down.

Shared Leads Create a Race. Not a Relationship.

The vendor sold the same contact to you and the school across town.

The student didn’t know that.

Your counselor is now competing on who dials first — not on who’s the better fit.

That’s a call-center contest. It rewards speed and volume. It punishes the campus that wants a real conversation about aid, placement, or whether the program is even right for this kid.

And families feel the difference.

One form. Five schools. A week of identical scripts. The student who wanted FAFSA help got a queue.

The school that looks familiar later? It’s the one they saw in a news story. A faculty quote. A piece about outcomes.

Not the one that left the third voicemail.

Families Decide on Value. Ads Don’t Get to Make That Case.

Hanover’s 2026 admitted-student survey is blunt: students still see value in a degree — but they won’t enroll if the price is opaque or the institution looks shaky.

Carnegie found the same thing in 2025: scholarships and aid topped the list of selection factors. Job placement, prestige, and whether the place feels real are climbing with them.

But EAB’s research is the part marketing teams skip. When students search online:

79% go to the school’s own site
29% click organic, non-ad results
8% click ads

Let me say that again. Eight percent.

They rate in-person events and college websites as trustworthy. Social posts and AI chatbots? Bottom of the list. And yet — a large share still use the chatbot anyway.

That gap matters. They’ll ask the model. They won’t trust the model. They trust journalism. Campus visits. The .edu.

Your paid lead never enters that trust stack. It enters the CRM.

The Desk and the Family Read the Same Things

Education reporters don’t file viewbooks.

They file enrollment numbers. FAFSA stories. Research that belongs in the public. Tuition math. Whether a program actually moved an outcome.

That is also what a parent forwards in a group text.

A ranking press release doesn’t do this. A “most innovative campus” roundup doesn’t do this. A student-newspaper clip sold as national coverage definitely doesn’t do this.

What files — and what a family will believe:

Who applied. Who enrolled. What broke in the process. A number, not a slogan about access.
A finding with a named faculty member and a method on the table.
Completion, placement, time-to-skill. What the campus saw after the pilot — not the demo.
A president, provost, or dean who will say the sticker price, the net price, and the trade.
Original alumni numbers: debt, jobs, who came back.

That’s earned media. A journalist chose to use you — by name — in a piece they actually filed. The URL is live. The quote can be checked.

A purchased lead can’t be checked.

It can only be dialed.

AI Already Sits Between Your Prospect and Your .Edu

Students now ask ChatGPT, Gemini, and Claude which programs are worth the debt. Which campuses are stable. What graduates actually earn.

Muck Rack’s May 2026 analysis of 25+ million cited links found:

84% of citations came from earned media
27% from journalism alone
0.3% from paid or advertorial content

The model is not reading your Meta lead ad.

It’s assembling answers from journalism, academic sources, government data, and third-party pages with a record.

If your campus is absent from the pieces those systems read, you’re absent from the shortlist a student sees at 11 p.m. — before they ever touch your inquiry form.

This isn’t a brand-awareness argument. It’s a discovery argument.

Search used to be ten blue links you could buy your way into. The answer box now prefers sources a desk already trusted.

The Cliff Doesn’t Care About Your CPL

WICHE projects college enrollments down 13% by 2041. Fall 2026 is the first class shaped by the 2008 birth-rate drop. High school graduates are projected to decline in 38 states.

Since 2016, more than 120 colleges and universities have closed or merged. At least 16 nonprofit institutions announced closures in 2025 alone. Roughly 1,100 of 1,900 colleges in one dataset had enrollment drops in three of five years. About 1,500 saw tuition revenue fall.

A lead buy can fill a weekly report while the institution shrinks.

That is how a campus dies with a healthy-looking dashboard.

The sector keeps diagnosing a capture problem. For most regional and small private institutions, the problem is demand. Fewer 18-year-olds. More skepticism about price. More questions about whether the place will still be open in four years.

You cannot purchase your way out of a smaller cohort with names four other schools also purchased.

You can become the campus a reporter names when the story is value, outcomes, or research.

That’s how demand shows up before the form.

Earned Media Is Not a Press Release About Being Listed

If the plan is a ranking play and a recap on your newsroom page, stop. Education desks have seen that packet. A hundred times.

Real earned media for a university — or an edtech product campuses actually use — looks like this:

Make the story first. An original number. A public finding. A person who will go on the record.
Work the book. Pitch the desks that already cover enrollment, aid, research, labor outcomes.
Collect live URLs. Named in the body. Not a logo in a roundup. Not a mention with no link.

That corpus does three jobs a lead vendor can’t:

It gives admissions a piece to send that wasn’t written by marketing.
It gives your .edu pages something to sit under besides a slogan.
It gives AI systems and journalists a source to reuse the next time the beat comes back around.

And media drives more media. One filed story is how the next producer finds you. The second desk didn’t get a purchased inquiry.

They got a clip.

Run the Comparison in Dollars. Not Vibes.

Take a round number.

You buy 500 shared leads at $180. That’s $90,000.

Half never respond → 250 contacts. At a 2% enroll rate → 10 students.

Cost per start: $9,000 — before counselor time, CRM costs, and the ones who melt over the summer.

Now put the same $90,000 into a campaign that produces a floor of real placements — named, in-content, live URL — on desks families and models already read.

You don’t get 500 rows in Salesforce on day three.

You get stories that keep working. In search. In AI answers. In counselor packets. In the next reporter’s file.

One is a consumable. The other is an asset.

What to Do This Month

Stop asking marketing for more leads. Ask for a story a desk can file.

Bring one of these to the table — with a person attached:

FAFSA and enrollment: who applied, who enrolled, where the process broke, what you changed
Net price, said out loud, by a named leader, with the trade
Placement and debt for a specific program — not the viewbook average
A research finding with method, sample, and a faculty member who’ll take the second question
An edtech pilot with completion or time-to-skill from a campus that will be named

Can’t name the person or share the number? You’re not ready for earned media.

You’re ready for another vendor deck.

Then measure the work the way finance respects:

Live URLs, not “impressions”
Name in the body, not a seal in a list
Referral traffic and assisted applications from those URLs
Whether the same stories show up when you ask an AI model the question a parent asks

Keep the CRM. Just stop treating a purchased row as proof the market wants you.

The Floor

PressTitan works the higher-ed beat the way the desk does. Enrollment. Research. Outcomes. Universities and edtech on one page. A named person in a piece a journalist actually filed.

5, 10, or 20 placements in 90 days. Or your money back.

Start here: presstitan.com/industries/higher-education