More than 30,000 leads a month — and no way to tell which ones would actually enroll. Every lead got the same treatment, the same resources, the same cost. Here is how one national online university fixed that with predictive scoring — and cut their marketing spend nearly in half.
"We were treating every lead like it was worth the same. Once we could see the ones that actually enrolled, we realized we had been spending half our budget chasing people who were never going to pick up the phone. The scoring model did not just save money — it completely changed how we work."
— VP of Enrollment Marketing, National Online UniversityLead volume without lead intelligence is just noise
Online higher education is one of the highest-volume, most competitive lead generation markets in existence. Universities spend aggressively on third-party lead channels — partner networks, aggregator sites, and paid media — to fill their enrollment funnels. The challenge is that lead quality varies enormously by source, and most institutions lack the analytical infrastructure to distinguish high-converting leads from low-converting ones in real time.
The result: enrollment teams work harder and harder as lead volume grows, without necessarily enrolling more students. Budget scales with volume rather than with outcomes.
30,000 leads a month — treated as equally valuable
This national online university was receiving more than 30,000 leads per month from a network of third-party partners. Operationally, every lead was treated equally — the same outreach sequence, the same call cadence, the same follow-up investment. The enrollment team was overwhelmed, and conversion rates were lower than they should have been.
Budget allocated to low-quality channels was subsidizing the waste without anyone having visibility into which leads were actually converting. The university was paying premium prices to pursue a large percentage of leads with near-zero enrollment probability.
Channel-level scoring helped — but only at the surface
The university had previously attempted a manual channel-scoring exercise — ranking their lead partners by historical conversion rate and adjusting spend accordingly. This helped at the channel level but did nothing to differentiate leads within a channel. A low-quality lead from a high-converting channel still consumed the same enrollment team resources as a high-quality one.
A scoring model built on their actual enrollment data
Versium began with a deep analysis of the university's historical lead data — enriching every past record with demographic, behavioral, and lifestyle attributes, then building a statistical model based on the observable differences between leads who enrolled and leads who did not. The model identified which combination of attributes most reliably predicted enrollment probability.
That model was deployed in real time against incoming leads: as new prospects entered the university's intake system, Versium scored each one automatically. High-score leads were routed for immediate, high-touch outreach. Low-score leads received automated, low-cost nurture sequences or were deprioritized entirely.
Nearly half the spend — same enrollment numbers
- Marketing spend reduced by 49% as budget was reallocated away from low-propensity lead sources
- Enrollment numbers held steady despite the significant spend reduction
- Enrollment team productivity increased as reps focused exclusively on high-probability leads
- Real-time scoring integrated directly into the lead intake system with no manual review required
- Low-converting partner channels renegotiated or defunded based on score distribution data
- Cost per enrolled student declined substantially as conversion efficiency improved
Intelligence compounds downstream
Volume without intelligence is not a pipeline — it is noise. This university discovered that nearly half of their marketing budget was being consumed by leads with almost no chance of converting. Versium's predictive scoring did not just save money; it gave the enrollment team the clarity to do their jobs better. That is the compounding value of data intelligence: it does not just cut costs, it makes every downstream activity more effective.
Are half your leads worth pursuing?
Versium REACH builds a predictive scoring model from your actual conversion data — so your team focuses on the leads that enroll, apply, or buy, not the ones that ghost.
Start your free trial