A public health sciences university in the Midwest trains pharmacists, clinicians, and biomedical researchers across a portfolio of graduate and professional degrees. Its enrollment team competes for two very different kinds of applicants at once: traditional students weighing their next academic step, and working healthcare professionals considering a return to school mid-career.
Propellant Media was awarded the account through a competitive RFP process in June 2025, against a field that included national higher-education marketing specialists. The proposal did not lead with a promise of more leads. It led with full-funnel attribution in a shared dashboard, and a commitment to feed the university’s own back-end enrollment data into forward enrollment forecasting.
The Challenge
Five graduate programs needed to go to market inside a single enrollment cycle, on a single budget that could not be spread thin. The portfolio ranged from an established pharmacy doctorate with years of historical lead data to a brand-new clinical mental health master’s with no baseline at all, plus three specialized research and innovation degrees with genuinely small addressable audiences.
Splitting the budget five ways meant every program had to earn its own share. None could be carried by the others. And the two audiences shared almost nothing — a 22-year-old chemistry major and a 38-year-old hospital clinician do not respond to the same message, the same channel, or the same call to action.
What Our Team Was Tasked With
Build awareness and generate qualified, trackable interest ahead of the fall enrollment cycle across all five programs, reaching both audience tracks, on a $38,000 gross campaign budget across a five-month flight. Propellant handled media strategy, geofencing, audience construction, reporting, and the full creative strategy and production workload.
Media Plan We Developed
- Geofencing: Two distinct fence sets per program. Traditional fences covered psychology, biology, and chemistry departments at colleges and universities across the region. Clinical fences covered hospitals, medical centers, and research institutions, reaching working professionals during the workday when a career-advancing degree is genuinely on their mind.
- Programmatic Display: Static banner inventory across news, weather, and lifestyle properties, carrying top-of-funnel awareness for each program against geofenced and in-market audiences.
- Meta (Facebook & Instagram): Every device identified inside a fence passed into Meta as a dedicated named audience, split by track and by program. Mid-funnel consideration creative held attention; bottom-funnel creative carried an in-platform lead form.
- Past-Enrollee Lookalikes: Built from the university’s CRM lists where a program had enough history to support one.
- In-Market Data Segments: For newer programs with no lead list to model from, behavioral segments signalling active interest in that specific field of study.
- Retargeting: Applied across both programmatic and Meta, closing the loop on anyone who had already engaged.
Creative Strategy And Production
Propellant wrote and produced the entire creative set rather than adapting client-supplied assets. Each program received four dedicated units: two top-of-funnel prospecting ads — one written for traditional students, one for working clinicians — a mid-funnel consideration unit built to hold attention, and a bottom-funnel conversion ad constructed around that program’s own lead form. Twenty units across five programs.
Messaging stayed short and emotionally direct, speaking to where each audience already was rather than opening with a list of program features. Four additional units shipped mid-flight when one program was split into two concentrations.
Campaign Results
| Program | Impressions | Clicks | Leads | Cost Per Lead |
|---|---|---|---|---|
| Pharmacy doctorate | 834,919 | 6,729 | 532 | $9.94 |
| Clinical mental health master’s | 735,607 | 6,299 | 369 | $14.18 |
| Health innovation master’s | 484,247 | 2,828 | 168 | $29.95 |
| Biomedical research PhD/MS | 454,909 | 2,651 | 215 | $18.55 |
| Health data science master’s | 385,905 | 2,306 | 134 | $29.59 |
| Total | 2,895,587 | 20,813 | 1,418 | $16.58 avg. |
Leads reflect Facebook and Instagram form submissions only. The university’s website inquiry forms were not trackable back to a specific ad during this cycle, so the true figure is higher than the one shown. Cost per lead reflects Meta spend against Meta-tracked leads.
| Biomedical research PhD/MS | 430% | |
| Health data science master’s | 268% | |
| Pharmacy doctorate | 152% | |
| Health innovation master’s | 112% |
What The Data Settled
Geofencing outperformed the client’s own CRM list. Half of all 1,418 leads — 712 completed forms — came from audiences first identified at a physical location and then worked through Meta to a submitted form. On the pharmacy doctorate the geofenced audiences returned 300 of 532 leads, or 56%, finishing ahead of both the past-enrollee lookalikes built from the university’s CRM and site retargeting.
The smallest budgets produced the biggest overshoots. The biomedical research PhD/MS cleared its goal by 430% and the health data science master’s by 268%, both on the smallest allocations in the set. That is a signal of real, previously untapped demand for specialized degrees once the targeting actually finds the right people.
A performance dip was a creative problem, not a media problem. When the health innovation master’s softened a few weeks in, the diagnosis was one generic message serving two distinct concentrations. Splitting the creative into two concentration-specific tracks fixed it, and the original general ad was retired once the new versions were live and performing.
In-state geofences did the heavy lifting. The university’s home state drove the largest share of both impressions and leads, and in three of the five programs, leads generated in-state cost less than the program’s overall average — exactly what a regional institution should want to see.
What We Take To The Next Campaign
- Geofence the building, not the city. Clinical and campus fences fed Meta a warmer audience than a CRM lookalike did. The precision of the fence, not the size of the geography, drove the result.
- Run an audience health check at three to four weeks. The mid-flight concentration split worked, but a scheduled early checkpoint on every program would have caught the mismatch sooner.
- Fix attribution before scaling spend. With only in-platform form submissions trackable, 1,418 is a floor rather than a ceiling. Conversion tracking on the university’s own inquiry forms is the first item of the next scope.
- Fund the overperformers. Programs that cleared their goals by 2.5x to 4x on the smallest budgets are the clearest candidates for incremental investment next cycle.
The Outcome
Every program with a lead goal cleared it, and the fifth — brand new, with no baseline to measure against — delivered 369 leads on its own terms. The engagement has since been renewed and expanded, growing from five programs to a planned eight or nine across the next two enrollment cycles, with website conversion tracking leading the new scope.
The relationship ran on a biweekly cadence with the full enrollment team rather than a monthly report drop. Across the entire cycle there was never a reset call — no point at which the work had to be re-explained or restarted. That, as much as the cost per lead, is what the renewal was built on.
Frequently Asked Questions
What is a good cost per lead for graduate program enrollment marketing?
This campaign averaged $16.58 per lead across five graduate programs, with the strongest program at $9.94 and the most specialized at $29.95. For graduate and professional healthcare degrees, anything under roughly $60 to $70 per lead is generally a healthy outcome, so this set performed well. The wider point is that cost per lead should always be read alongside volume and audience size — a highly specialized research degree with a small addressable audience will cost more per lead than a pharmacy doctorate, and that is a function of the market rather than a failure of the campaign.
Does geofencing actually generate enrollment leads, or just awareness?
It generates leads when the geofenced audience is carried forward rather than abandoned after a single impression. In this campaign, 712 of 1,418 completed lead forms — a full 50% — came from audiences first identified inside a geofence. The mechanism matters: devices identified at target campuses, hospitals, and research institutions were passed into Meta as dedicated named audiences per program, then worked through consideration and conversion creative until the person submitted a form. Geofencing on its own builds awareness. Geofencing feeding a social retargeting funnel builds a pipeline.
How do you market a brand-new degree program with no historical data?
You replace historical lead data with behavioral and physical intent signals, and you resist setting a fabricated goal. The new clinical mental health master’s in this campaign had no prior-year lead volume, so no meaningful lead goal could be set for it. Instead of inventing one, the program was targeted through in-market data segments signalling active interest in the field and through clinical geofences reaching people already working in adjacent roles. It delivered 369 leads at $14.18 — the second-highest volume in the set — measured against applicant and enrollment goals rather than a made-up lead target.
Why does website conversion tracking matter for enrollment campaigns?
Without it, you can only count the leads that happen inside the ad platform, which understates the campaign. In this cycle only Facebook and Instagram form submissions were attributable, meaning any prospective student who clicked an ad, visited a program page, and submitted an inquiry through the site was invisible to reporting. The 1,418 figure is therefore a floor rather than a total. Adding conversion tracking to institutional inquiry forms lets you attribute website leads to the specific ad and audience that produced them, and eventually to report cost per enrolled student rather than only cost per lead.