Measuring the retention ROI of a co-curricular program

Retention ROI arguments fail in finance review for predictable reasons, and none of them is that the program does not work. They fail because the denominator is undefined, the comparison group is the institutional average, or the cost figure omits the staff time that makes the program run.

Here is how to build the version that survives.

Step 1: define participation before the term starts

Write down what counts as participation while you still have no idea what the number will look like. Attendance at how many sessions? A completed milestone? A sustained role?

Choosing the threshold after seeing the data is the single most common flaw in co-curricular retention claims, and it is visible to anyone who has reviewed a few. A definition set in advance and held constant for three years is worth more than a favorable definition set in arrears.

Document it somewhere dated. An email to yourself is sufficient.

Step 2: build a comparison group, not a benchmark

Comparing participants to your institutional average overstates the effect, because participants opted in. This is the objection a competent provost or CFO raises first, and it ends the conversation if you have no answer.

The workable answer is a matched group drawn from your own student information system:

• Same entry cohort. Fall 2025 first-time students compared to fall 2025 first-time students.
• Similar credit load. Full-time and part-time students persist at different rates for reasons unrelated to your program.
• Similar Pell status. Financial pressure is the largest confounder in the data you have access to.
• Similar prior involvement, where you can capture it. If both groups had no prior co-curricular participation, you have removed most of the selection problem in one field.

You do not need a statistician for this. A 200-student participant list and a 200-student matched list pulled on those four fields is dramatically more credible than a comparison against a 71% institutional figure.

Step 3: cost the program honestly, staff time included

The line item is not the cost. The cost is the line item plus the hours.

For a like-for-like comparison, price three things:

• Direct institutional cost: what the program charges you, and on what schedule. Sometimes this is zero; confirm it rather than assuming either way.
• Staff hours per term, at loaded salary. This is where in-house programs quietly lose to external ones, because building curriculum, recruiting, tracking, and assessment are all staff hours that never appear in a budget request.
• Student cost, stated separately. If students pay anything, name it plainly in the proposal. It is one of the most common objections and it does not improve by being buried.

As a reference point on how these components separate: the NSLS charges the institution nothing, publishes an advisor commitment of 10 to 20 hours per semester, and has a separate one-time $95 enrollment fee paid by students who accept a nomination. Those are three distinct things and they belong in three distinct lines. Our page on what an NSLS chapter costs sets them out, and the advisor time commitment covers the hours.

Step 4: pick a denominator for the value side

Retained students have a value your institution can already compute, and finance almost certainly has a figure. Ask for theirs rather than inventing one: net tuition revenue per retained student per term is the usual form, and using their number removes an entire category of argument.

Then be conservative on the numerator. If your matched comparison shows a five-point difference, model three points. A proposal that survives a hostile reading at three points is stronger than one that requires five.

Step 5: report reach separately, because it needs no causal claim

The most defensible finding in most co-curricular programs is not a retention lift. It is reach.

“We engaged 210 commuter students who had no prior co-curricular involvement, and 140 of them returned the following term” is a real result, requires no comparison group, and is very hard to argue with. Present it alongside the ROI model rather than inside it, so a challenge to the causal claim does not take the reach finding down with it.

Disaggregate it. An overall figure carried entirely by residential full-time students is a different finding than it appears, and discovering that in the meeting is worse than discovering it beforehand.

What published outcomes are good for

Named institutional outcomes are useful for establishing that the mechanism is real, not for sizing your effect. NSLS chapters have reported an 81% semester-to-semester retention rate among inducted students at Delgado Community College (nearly 20 percent higher than that institution’s overall rate), 95% retention among students who joined the NSLS at Merrimack College, and, at St. Petersburg College, a retention rate 37 percent above the college-wide rate among 24 members of a one-campus pilot.

They measure different things. Delgado’s and St. Petersburg’s are stated against the institution’s overall rate, St. Petersburg’s for 24 students in a one-campus pilot; Merrimack’s is a rate for members alone. In each case students chose to join, so selection is part of the result. Cite them as evidence that structured involvement is associated with persistence across institution types. Do not use them as your projection, and do not let anyone else use them as your projection either.

The number your model rolls up against

Median full-time retention across US degree-granting institutions is 71%, from 77% at private nonprofit four-year institutions to 63% at public two-year institutions (IPEDS, Fall 2023, EF2023D). Model against your sector median rather than the national figure. A four-point gain at a community college and a four-point gain at a selective private institution are not the same claim, and reviewers know it.

Do this next

Before you build anything, write one sentence defining participation and email it to yourself with today’s date. That single act removes the most common objection to co-curricular retention claims, costs nothing, and cannot be done retroactively.

For sector benchmarks and how to construct the comparison itself, see our guide to student retention and persistence.