For VPs of Enrollment & Admissions

See the colleges AI names instead of yours

We run 31 college searches the way a seventeen-year-old actually types them. 10 of them never mention your name. Whether you show up in those is the whole report.

  • ✳Your AI consideration set share — how many unbranded searches surface you, and the ranked list of schools that surfaced instead.
  • ✳What the models get wrong — tuition, acceptance rate and program claims, pulled out so you can check them against your own IPEDS submission.
  • ✳Whose pages they read — the domains cited when a model describes you. Usually not yours.
Run the audit

Three fields. We work out your metro, top programs and peer set from there.

Takes about three minutes. No email needed to see your score.

The method

Six buckets, 31 prompts, one institution

Written in student phrasing, not institutional vocabulary. Nobody types “undergraduate value proposition.” The wording is deliberate — cleaning it up changes what comes back.

01 / COLD LIST-BUILDING

Unbranded searches a real prospect types before they know your name. Whether you appear here is the whole finding.

02 / PROGRAM-LEVEL DEMAND

Program invisibility is more actionable than institutional invisibility - and it is the finding a dean will act on.

03 / AFFORDABILITY AND AID

Where regional privates lose the most ground, and where wrong numbers do the most damage.

04 / CONSIDERATION SET

The schools a model volunteers alongside you. Frequently not your official peer set.

05 / REPUTATION AND ACCURACY

Tuition, acceptance rate, program list, framing. Verify every figure here against your IPEDS submission.

06 / DEAL-BREAKER FILTERS

Test-optional, rolling admissions, late availability - the constraints that decide a late-cycle application.

A few of the unbranded ones, verbatim

affordable private colleges near Philadelphia with strong nursing programs
colleges that give good merit scholarships to B students
schools like Villanova but easier to get into
small colleges where I won’t be just a number
01 / THE HEADLINE

Consideration set share

Appearances across the 10 unbranded searches, then the ranked table of schools that appeared instead of you. Most regional institutions score 0–2.

02 / THE ALARMING PART

Every checkable claim

Tuition, acceptance rate, aid and program assertions, flagged when the models contradict each other. This is the page that gets forwarded to a president.

03 / THE FIXABLE PART

Cited domains

Which pages a model reads to describe you. When Niche, Reddit and US News outrank your own site, that is a content problem with a known shape.

What this report does not claim. AI visibility is not why your applications moved last year. The demographic cliff, FAFSA disruption and the 2025 enrollment peak are real and better documented. Treat this as a leading indicator of where discovery is heading — every figure is timestamped, and labelled with the model and date that produced it, so a provost can argue with it fairly.

The honest handoff

This report diagnoses something we don’t sell a fix for

You cannot buy placement in an AI answer, and moving it takes twelve to eighteen months of content and structured-data work. What Kollegio does is different: the students running these exact searches are on our platform now, describing what they want in plain language — faith-based schools where they can play soccer, environmental science near the outdoors — and their profiles can land in your Slate instance in about two weeks. Different route to the same students.

Run your audit first →What Kollegio does