At forty-four percent of American colleges, whether you opened the email, showed up at the info session, or visited campus changes the odds. Across the elite tier, the practice is sharply, unevenly applied — and the schools that lean on it hardest are also the ones with the most to lose.
Tucked into the Common Data Set, on a sheet most applicants will never see, admissions offices answer a single question every fall. How important is the level of applicant's interest?
There are four boxes — Very Important, Important, Considered, Not Considered — and across the 55 schools we track, the distribution is surprisingly lopsided. Most of the elite tier checks the last box. A handful checks the third. Three check the second. None checks the first.
That headline number hides a steep ranking effect. The schools at the very top — the ones whose yields exceed seventy percent — almost universally ignore engagement signals. Below them, the practice creeps in.
Plot the same data against tier, and the cliff becomes obvious.
At HYPSM, the answer is unanimous: zero out of five list interest as anything but Not Considered. At Ivy+, only Dartmouth dissents. Move down the prestige ladder and the proportion flips.
Among Selective private universities — schools with seven to twenty percent admit rates but yield rates well below thirty percent — interest is suddenly weighted. Tulane and Northeastern, both Tier 4, are the only two universities in the 55 to mark interest as Important.
The public flagships are a clean group: every single state university in the set ignores DI, for reasons of equity, scale, or law (UT Austin's top-six-percent auto-admit makes it irrelevant).
A pattern is emerging. The schools that weigh interest are also the ones with low yield.
Yield — the share of admitted students who actually enroll — is the metric admissions deans guard most jealously. It feeds directly into U.S. News selectivity scores. It dictates how many to admit. It is, ultimately, what demonstrated interest is for.
Above seventy percent yield, no one tracks engagement. Harvard's eighty-four percent makes the question moot — almost everyone admitted shows up. Below twenty-five percent yield, the calculus reverses. Every admit who declines is a forecasting error, a tuition revenue miss, a ranking penalty.
Yield management is the discipline of admitting people who will say yes. Demonstrated interest is the cheapest signal we have. — Synthesized from enrollment-management literature
Theory meets evidence next: a single peer-reviewed study quantifies the effect.
The most rigorous quantitative evidence on demonstrated interest comes from a 2019 study by Howell and Patel in Contemporary Economic Policy. They obtained admissions records for 12,501 applicants across two cycles at a "medium-sized, highly selective university" — widely understood to be Lehigh, which lists DI as Important in CDS.
Controlling for SAT, GPA, and other applicant attributes, they found that an on-site contact — most often a campus visit — raised the probability of admission by roughly thirty percentage points. Off-site contacts, like college fairs and high-school visits, added ten to thirteen points. Together, the boost reached twenty-one to twenty-four.
Crucially, the effect compounds with academic strength. A top-quartile SAT applicant who visited campus saw a thirty-four-point lift. A second-quartile applicant saw twenty-two. The signal works because it converts an already-plausible candidate into a likely-to-enroll one.
But not all signals are equal. Some are cheap, some are expensive — and cost is the point.
The Lehigh paper's enduring contribution is its argument that signal cost matters. A campus visit is expensive — flights, lodging, school days missed. An email open is free. The schools that track interest weight the two accordingly.
Top of the hierarchy: the in-person visit. Below it, the on-campus interview, then regional info sessions and college fairs. At the bottom — barely registered — sit email opens, portal logins, and social-media follows. Forty-eight of the top fifty universities use Technolutions Slate CRM to track all of this, but only the costly signals move admissions decisions.
The implication for an applicant is straightforward. Time and money spent visiting a campus that takes DI seriously can move the dial. Logging into the portal three times in a week does almost nothing.
If yield protection has a single mascot, it is a school in New Orleans.
No school in our 55 leans on demonstrated interest like Tulane. Its public-facing acceptance rate hovers around nine percent. That number, however, is an average that conceals two very different worlds.
Apply Early Decision, and the rate jumps to about sixty-seven percent. Apply Regular Decision, and you are looking at under three percent. The university fills the vast majority of its class through binding-commitment rounds — ED, EA, ED2 — leaving Regular Decision to function as a near-rejection pile.
Tulane's regular-decision rate is what statisticians would call functionally zero. A binding application is the only way in.
This is what aggressive yield management looks like in practice. Each round earlier, each campus visit logged, each email opened tilts the model further. Our simulation assigns Tulane a logit boost of +0.40 — the highest in the set, twice that of any other school.
Tulane is the loud case. Yield protection has quieter cousins.
"Tufts Syndrome" — the term, named for the Massachusetts university — is shorthand for rejecting or waitlisting applicants whose stats far exceed the school's median. The reasoning is statistical: those students will likely choose somewhere more selective. Better to waitlist than to admit and miss.
Four of our 55 schools carry the yield-protection label in industry databases: Tulane, Northeastern, Tufts, and UChicago. Tufts has somewhat moderated. WashU, the fifth historical name, dropped DI entirely in 2022, citing equity concerns for rural and first-generation applicants.
Yield protection rarely operates alone. It travels with heavy ED reliance, Slate-driven engagement scoring, and a "Why this school?" supplemental essay that functions as a written DI signal.
All of which suggests a final mechanical question. How much, in raw probability terms, can engagement actually shift a borderline application?
To translate research into a simulation, the Lehigh effect maps to a logit-space additive term. At a school that calls DI Important and an applicant with maximum demonstrated interest, the boost is roughly +0.4 in logit units — about a thirty-percentage-point swing at a 35% baseline admit rate.
At Considered schools, the boost is one-third to one-half as large. At Not Considered schools, it is zero. And the boost is heavily down-weighted in early rounds: applying Early Decision is itself the maximum DI signal — a binding commitment renders portal logins redundant.
The simulation applies a round multiplier: 0.2 for ED, 0.5 for EA, and 1.0 for Regular Decision. Demonstrated interest, in other words, exists primarily to triage RD applicants who passed on the binding option.
That asymmetry helps explain why disadvantaged applicants — who cannot afford the campus visit, the SAT prep, the binding ED commitment without seeing aid offers — are doubly penalized at DI-tracking schools. Several of those schools have started to notice.
Demonstrated interest is, in the end, a middle-of-the-distribution practice. The most selective schools do not need it; their pools are self-selected and their yields are insulated. The least selective schools cannot enforce it; their applicant pools are too transactional. Between those poles sit the enrollment-managed selectives — Tulane, Northeastern, Wake Forest, Tufts — where the calculus of yield protection makes every campus visit a small, real thumb on the scale.
For applicants, the practical advice is narrow but durable. If a school you care about lists DI as Considered or Important on its CDS, visit. Open the emails. Apply Early Decision if you are confident. If a school says it does not consider DI, take the school at its word — your time is better spent on the application itself, the essays, the activities list.
For the system overall, the ethical questions are sharpening. A signal that requires travel and time off-school is an access tax. WashU dropped DI in 2022 explicitly to remove that tax. Whether others follow will depend on the rankings calculus and the tolerance of admissions deans for one more annual yield surprise.