How we choose recommended schools

Every figure on a recommendation card comes from public federal data:

• U.S. Department of Education, College Scorecard for graduation rates, first-year retention, acceptance rates, loan repayment, median debt, cost of attendance, and earnings after graduation.
• IPEDS Student Financial Aid for net price broken out by family income band. This is the price we compare on whenever we know your income range.
• Census Post-Secondary Employment Outcomes (PSEO) for earnings by institution and field of study.
• Regional tuition exchange programs such as the Western Undergraduate Exchange, for cases where an out-of-state school charges close to in-state rates.

Where a school has not reported a figure, or the federal source suppresses it for privacy, we say so on the card rather than estimating it.

We do not narrow the list to your state or your region. The pool starts as every institution in the country, and a school's location matters only through what it does to your price: whether you would pay resident rates, whether a tuition exchange program applies, and how generous the school is to students from out of state.

If you want to limit the search geographically, that is your call to make. The distance control on the recommendation section is set to "Anywhere in the US" by default, and you can change it at any time. We never infer it from your address.

Before any school can be recommended to you, it has to clear these checks:

• Same kind of school. A two-year college is compared with two-year colleges, a four-year with four-year institutions. We do not mix them.
• Realistic to get into. Rather than predicting your admissions odds, we look at the schools you have already added and stay within a similar selectivity range, widened by a margin. It is a deliberately cautious approach: it avoids showing you schools far outside the range you are already considering, without pretending to know whether you would be admitted.
• Enough published data. A school with no usable price or outcome data is not recommended at all.
• A graduation-rate floor. See below. This one matters most.

A school is never recommended to you just for being cheap. Institutions below a minimum graduation rate are excluded from every recommendation type, no matter how large the cost advantage.

We set that floor from the actual national distribution rather than picking a round number. For four-year institutions and certificate-focused schools the floor is a 25% completion rate. For community colleges it is 20%, because applying the four-year number to two-year institutions would exclude roughly a fifth of them, including many open-access colleges doing genuinely valuable work with students who arrive with the greatest need. A school that reports no graduation rate at all does not clear the floor either, because we cannot verify it.

This is the guardrail that keeps the recommender honest. The Worth-It Score exists partly to expose schools that are inexpensive but do not graduate their students, and a recommendation engine that quietly pushed you toward those schools would undo that.

Every card carries a label explaining why that school appeared. A school that qualifies under several labels is shown under the strongest one, so the list does not repeat itself.

If a recommended school has a lower graduation rate than the school you are considering, that appears on the card. Recommendations that hide the tradeoffs are advertisements, not recommendations.

We also never tell you a recommended school is better than the one you picked. You may have very good reasons for your choice that no federal dataset captures. The point of this section is to widen the set of options you are weighing, not to second-guess it.

Every card shows the specific numbers that produced it, so you can reconstruct why you were shown a school and disagree with it if you want to.

No. Nothing in this section is paid. Recommendations are generated entirely from the federal data described above, and no school, advertiser or partner can buy a place in them or influence their order.

If that ever changes, paid placements would be clearly labeled, kept separate from the organic list, held to the same quality floor, and shown with the same metrics including unfavorable ones. We would update this page before any such change went live, not after.

• Federal data lags. Graduation rates and earnings describe students who enrolled years ago, which is the best available evidence but not a forecast of your experience.
• Net price by income band is an average for students in that range at that school. Your actual award letter is the only figure that is truly yours.
• Earnings differences reflect who attends a school and what they study as much as what the school adds.
• Acceptance rates are reported by only about a third of institutions, and many that do not report are open-admission.