1 · Why blended averages mislead
The most common cost-per-lead calculation divides total paid media spend by total leads — paid, organic, referral, everything. It feels reasonable. It is also the single most common way enrollment marketing reports flatter themselves.
Here's the failure with made-up but realistic numbers. Suppose you spend $100,000 and acquire 4,000 leads — 2,500 from paid channels and 1,500 from organic search and referrals. Blended CPL: $25. Paid-source CPL: $40. The blended number is 37% lower, and not one dollar of the spend bought those 1,500 organic leads.
Every decision made on the blended number inherits the distortion. Paid channels look cheaper than they are, so they keep their budgets. And when organic volume grows — a good thing! — paid channels look like they're improving even when they aren't.
2 · CPL and cost-per-enrollment done right
The honest definitions are simple:
CPL = paid media spend ÷ paid-source leads
CPE = paid media spend ÷ paid-source enrollments
The denominator only counts leads and enrollments attributed to channels you pay for. Organic, referral, and walk-in volume is reported — proudly — but it never enters the cost math, because no spend produced it.
Do this by channel, not just in total. A blended paid CPL across search, social, and display hides the same way a blended paid/organic CPL does: your best channel subsidizes your worst in the average. The decision-grade number is CPL and CPE per channel, per program.
Want to see your own gap? The honest cost calculator computes both versions from your numbers, in your browser.
3 · Gross vs. net spend
If an agency runs your media, there are two true spend numbers: what the platforms charged (net) and what you paid including agency fees (gross). Both are legitimate — they answer different questions. Net spend evaluates the channel; gross spend evaluates the investment.
The mistake is mixing them silently. A report that compares this year's gross to last year's net will invent a trend that doesn't exist. Pick the lens per question, label it, and keep per-channel fee markups explicit so anyone can reconcile the two.
4 · Period vs. cohort: two questions, two views
“Is this channel working?” has two honest answers depending on what you're deciding.
Period view counts everything that happened between two dates: leads in, applications in, enrollments in, spend out. It's the right lens for monthly reporting and budget reviews. Its blind spot: the enrollments it counts mostly came from leads acquired before the period, and the leads it counts mostly haven't converted yet.
Cohort view takes the leads acquired in a window and follows them forward to today, however long conversion takes. It's the right lens for judging a campaign: these are the leads it bought, and this is what they eventually did. Young cohorts are incomplete by definition — which is why a cohort view needs maturity curves, so you can compare a six-week-old cohort to how past cohorts looked at six weeks.
Most reporting only gives you the period lens. The most expensive mistakes — killing a campaign whose leads hadn't matured, or funding one whose early volume never converts — live exactly in that gap.
5 · Deciding in-cycle: pacing
A perfect post-mortem is worth less than a decent mid-cycle read, because only one of them can still change the outcome. Pacing is the discipline of comparing three clocks continuously: budget consumed, cycle elapsed, and funnel produced.
A channel that has spent 80% of its budget at 60% of the cycle isn't necessarily failing — but it's a conversation to have now, with reallocation still on the table. The pattern to avoid is structural: if performance data arrives monthly and decisions happen quarterly, every course-correction lands a season late.
6 · The checklist
Six questions to ask of any enrollment marketing report:
- Does CPL divide spend by paid-source leads only?
- Is cost per enrolled student computed at all — and per channel?
- Are gross and net spend labeled, with fee markups explicit?
- Can you see a cohort view, or only date-range totals?
- Is pacing compared against both budget and cycle progress?
- How old is the data — and would you know if a sync broke?
Pennant answers yes to all six out of the box — that's what it was built for. If your current reporting already does too, genuinely: keep it.