Retention for commuter and online college students

A commuter student is on campus for ninety minutes and then in a car. A fully online student is never on campus at all. Both are enrolled, both are counted in your retention denominator, and neither is reachable by the programming that most institutions run.

This is not an engagement-effort problem. It is a structural mismatch, and it is fixable with four design changes.

Why these students leave

The attrition mechanism is the same one that operates everywhere: attrition is cumulative, and a student with no attachment leaves over something small because nothing is holding them. What differs for commuters and online students is that the institution has almost no surface area to attach to.

Consider what a residential student accumulates passively: a roommate, a floor, a dining hall, incidental contact with staff, and eight unstructured evening hours a week within walking distance of every program the campus runs. A commuter accumulates none of it. An online student accumulates none of it and has no physical referent for the institution at all.

The sector data reflects where this concentrates. Median full-time retention across US degree-granting institutions is 71%, but the median at public two-year institutions is 63% against 77% at private nonprofit four-year institutions, a 14-point gap (IPEDS, Fall 2023, EF2023D). Community colleges enroll a far higher share of commuting, part-time, and working adult students. The structural gap and the population list are the same finding.

Change one: stop scheduling in the evening

Evening programming is a screening mechanism. It selects for students who live on or near campus, do not work a shift, and have no dependents, which is to say, it selects against the population you are trying to retain.

The alternatives are unglamorous and effective:

• Midday, between class blocks. A commuter’s only genuinely available hour is the one they are already on campus for.
• Immediately before or after a required course, in a room near it. Distance and dead time are the actual barriers, not interest.
• Asynchronous by default, with live sessions as an option rather than the spine.

Change two: design for online students rather than adapting

A virtual version of a residential program is worse than either. It carries the assumptions of physical presence (a fixed hour, a shared institutional rhythm, incidental social contact) without the infrastructure that made those assumptions reasonable.

Programming that works for online students has different properties. It lives where the student already is, usually a course environment or a group thread rather than an events calendar. It is completable across weeks rather than attended at an hour. And it produces something portable (a credential, a project, a résumé line) because the campus wall means nothing to a student who will never see it.

Delivery mode is worth checking explicitly when you evaluate any external program. The NSLS, for instance, is available in-person, virtual, and fully asynchronous, and runs a National Online Chapter for institutions without a campus chapter. Whether a program has a genuinely asynchronous pathway, as opposed to a recorded one, is the question to ask.

Change three: use peer groups as the attachment point

Commuter and online students cannot be retained through staff contact at volume. There are not enough staff hours, and the students are not present when the hours exist.

Small peer groups solve this because they operate on the student’s schedule rather than the institution’s. A group of six with a standing commitment and a shared goal notices an absence faster than any alert system and reaches the missing member more credibly than an office does.

Two details matter. The group must be fixed: the same people, so absence is legible. And it must have a purpose beyond socializing, because groups organized around belonging dissolve while groups organized around a task produce belonging as a byproduct. Structured peer formats such as small goal-setting teams work for this reason: the task carries the group, and the group carries the student.

Change four: recruit by nomination, not announcement

A campus-wide email reaches students who read campus email, which correlates with students who are already attached. The commuter who checks their portal for grades and nothing else does not see it.

Nomination inverts the filter. When a faculty member or advisor names a student, the students named are frequently the ones who would never have raised their hand. For commuters and online students, a course instructor is often the institution’s only real point of contact. Build the invitation into that relationship rather than into a marketing channel.

What to measure

Do not report an overall engagement figure. Report it disaggregated, every time:

• Distinct commuter students engaged, as a share of commuter enrollment.
• Distinct online students engaged, as a share of exclusively-distance enrollment. IPEDS publishes an exclusively-distance share per institution, so your denominator is defensible.
• Part-time engagement rate, separately from full-time.
• Repeat engagement term over term within each group, which is the closest available proxy for whether anything stuck.

An overall lift concentrated in residential full-time students is not the finding it appears to be, and disaggregating before the meeting rather than during it is the difference between a program renewal and a program review.

The pressure underneath

Financial stress is the constant in this population: 59% of college students nationally have considered dropping out because of it (Ellucian, 2024). Working students and commuters carry more of that load and have fewer buffers.

An engagement program does not solve it. What it changes is whether a student under that pressure has anyone who notices, and whether they know where the door to help is. That is a modest claim and it is the honest one.

Do this next

Pull your engagement participation list and join it to two fields you already have: commuter status and delivery mode. If either group’s participation rate is materially below its enrollment share, you have located your retention gap and you did not need a new system to find it.

For sector benchmarks and how to build a comparison leadership will accept, see our guide to student retention and persistence. For the design side, see student engagement and belonging in higher education. Each population also has its own page: commuter campuses and online and adult learners.