The current sense of enrollment managers is that the present is hard, and the future is going to be even harder. In light of that, it’s natural to look in different directions to bolster enrollment, and it’s generally not long before someone points out that “it’s cheaper to retain a student than it is to recruit a new one.” My former boss, David Kalsbeek (whom I’ll cite frequently here) will point out that this statement doesn’t mean what people think it does: If you are working with a single student, and you determine a $2,500 increase in institutional aid will make the difference between a student leaving and staying, then, yes, it’s true.

But there are few–if any–institutions that can operationalize that approach, let alone make accurate assessments of a student’s propensity to stay or leave. So if you try to do that, it’s likely you will spend considerably more to retain students than to recruit them; your solutions are almost always applied with a 4″ painter’s brush, not an artist’s brush.

This all ignores, of course, the fact that retention and graduation–while not quite destiny–are largely determined by the students you admit. Not convinced? If you do a simple regression comparing median SAT scores with six-year graduation rates, this is what you get.

The SAT is a good indicator because it rolls up several variables that contribute to college graduation: Academic performance, ethnicity, parental attainment, and perhaps most important, family income. (If you believe SAT and GPA are the sole factors that go into academic success, we really need to talk.) Institutions at the top right of this distribution take very few risks in the admissions process, and they have the ability to offer extraordinary funding to those who have financial challenges, and that translates into higher graduation rates.

You should notice, however, that at any point along the X-axis, there is some considerable variability, and that’s your opportunity. At the same time, you have to understand that there are thousands of interactions and variables we will never be able to collect or analyze, and trying to figure out how to surgically address your challenges is likely to be a missed opportunity.

Whatever you do, invest in things that help all students (such as advising, career services, or curricular change), and avoid the tendency to create programs or approaches that focus on small subsets of students. Many of the students who leave aren’t in your high risk groups; and many of the students in the high risk groups are already graduating. In short, doing better for everyone is almost certainly the best and the most efficient way to increase your retention and graduation rates.