The dominant narrative in higher-ed marketing is built on a single assumption: selectivity sells.
It does, in 33 places. Everywhere else, it's a trap.
What the data says
From the U.S. Department of Education's College Scorecard 03/2026 release, four-year U.S. institutions only:
In fourteen years, the typical American four-year college went from admitting two-thirds of applicants to admitting more than three-quarters. The COVID cycle accelerated it. It has not unwound.
One in five U.S. four-year colleges admits more than 90% of applicants. Marketing themselves as selective is not just inaccurate — for any prospective family that has spent thirty seconds on the Common Data Set, it is actively brand-damaging.
The earnings curve flattens at the 25% line
Now look at admit rate against the median 10-year post-entry earnings of graduates:
| Admit rate | N | Earnings 10y | Median grad debt | 6-yr completion | Median Pell % |
|---|---|---|---|---|---|
| <10% | 33 | $92,192 | $14,000 | 94.1% | 17.9% |
| 10–25% | 53 | $77,234 | $18,625 | 89.7% | 16.2% |
| 25–50% | 163 | $57,103 | $23,590 | 62.5% | 27.9% |
| 50–75% | 455 | $54,396 | $23,644 | 57.5% | 32.0% |
| >75% | 824 | $54,106 | $23,592 | 54.5% | 34.2% |
The earnings gradient is non-linear. It crashes at the 25% admit-rate line, then is essentially flat for the next 1,442 institutions.
From the prospective student's wallet perspective, the school admitting 30% and the school admitting 95% will produce nearly identical median outcomes.
Selectivity is a winning brand for 86 institutions. For the other 1,707, it is the substitute for the work that would actually move enrollment.
The variance within tier matters more than between tier
Compare the 25th-percentile and 75th-percentile graduate within each admit-rate band:
| Admit rate | Earn p25 | Earn 10y | Earn p75 | Within-tier spread |
|---|---|---|---|---|
| <10% | $54k | $92k | $147k | $93k |
| 10–25% | $48k | $77k | $124k | $76k |
| 25–50% | $37k | $57k | $86k | $49k |
| 50–75% | $35k | $54k | $79k | $44k |
| 75–90% | $35k | $54k | $78k | $43k |
| >90% | $34k | $54k | $77k | $43k |
A graduate from a sub-50%-admit school in the 25th percentile earns $37K. A graduate from a >90%-admit school in the 75th percentile earns $77K — more than double.
Which school you go to matters less than which student you are at that school, and which program you complete. That insight should reshape how every institution at >25% admit rate writes its program pages, structures its proof-of-outcomes content, and answers AI engines.
Why this matters for AEO
ChatGPT, Perplexity, AI Overviews, and Gemini increasingly answer the "is school X selective" question with concrete numbers. They pull from the Scorecard. They pull from the institution's own Common Data Set. They pull from peer-set tables. If your marketing copy asserts a level of selectivity the data doesn't support, the engines will publish the disagreement — visibly, in the answer.
We've watched this happen in real time. A regional state institution in our anonymized panel has a homepage hero claiming "highly selective." Their Scorecard admit rate is 87%. Every AI engine we tested correctly cited the admit rate from the Scorecard while quoting the homepage. The contradiction was visible to any prospective family inside thirty seconds of asking.
What to do instead
Three substitutions for the schools that can't truthfully market on selectivity:
1. Lead with completion and earnings, not admit rate
If you complete 65% of your students in six years against a national median of 56%, that is the headline. If your median graduate out-earns the BLS median associate's holder in your state by $12K, that is the headline. The numbers exist. Most institutional websites bury them.
2. Lead with program-level outcomes, not aggregate
The Scorecard publishes field-of-study earnings and debt at a program level. Your nursing program's median graduate earns $75K. Your computer science program's median graduate earns $84K. That is the page that ranks for "best [program] in [region]" in the AI surface — not the aggregate page that buries the numbers under brand copy.
3. Lead with fit, not prestige
The 1,707 non-selective four-year institutions in the U.S. are competing for the same students. The differentiator is no longer "we say no to most people." It is "you specifically belong here, and this is why." The "why" needs sources, not adjectives.
Bottom line
If your institution sits at >25% admit rate, you cannot win on selectivity. You also can't win on aggregate earnings — the variance there is small. You can only win on fit, on program-level outcomes, and on lived experience, communicated with enough specificity that families recognize themselves in your campaign.
That is the work. Selectivity theater is the substitute for it. The College Scorecard release of 03/2026 makes the substitute increasingly hard to maintain.