Edtech refunds and dropout economics, explained
Gross bookings flatter edtech unit economics. Refund windows, cancellation rates and dropout curves decide whether a cohort ever pays back its acquisition cost.
The short answer
Edtech refunds and dropout economics is the practice of measuring revenue only after refunds, cancellations and learner dropouts are removed. A short refund window, a high cancellation rate and a steep dropout curve all push back CAC payback. Founders should model net revenue per enrolled learner across the full course life, not gross bookings at sign-up.
Key takeaways
- Gross bookings overstate edtech health because refunds, cancellations and dropouts reduce the revenue a company actually keeps.
- Revenue recognition spreads a course fee across the delivery period, so a large upfront collection is deferred revenue, not earned income.
- CAC payback should be measured from the close of the refund window, using net revenue and contribution margin, not gross bookings.
- Cohort dropout curves are usually steep early, so average-retention assumptions overstate revenue in the months when CAC needs recovering.
- Model net revenue per enrolled learner month by month, with delivery costs included, before quoting any payback figure.
Edtech refunds and dropout economics is the discipline of counting only the money a learning company actually keeps. A learner enrols, pays a fee, and may leave inside a refund window, cancel a subscription, or simply stop attending. Each of those events reduces revenue. If a company reports gross bookings — the total value of enrolments sold — and ignores refunds and dropouts, its unit economics look far healthier than they are.
This explainer sets out how refund windows, cancellation rates, cohort dropout curves and revenue recognition interact, and why the right metric is net revenue per enrolled learner rather than gross bookings.
What is the difference between gross bookings and net revenue in edtech?
Gross bookings are the headline number: seats sold multiplied by price. Net revenue is what remains after refunds, cancellations, discounts and the portion of fees that must be recognised over time as the course is delivered. In Indian edtech, the gap between the two can be wide because many courses run for months and carry an explicit money-back period.
Revenue recognition is the accounting rule that decides when a fee becomes revenue. Under standard practice, a fee collected in advance is a liability — deferred revenue — until the service is delivered. If a learner pays for a 12-month programme, the company cannot book the full amount on day one. It recognises revenue month by month, or module by module, as classes are taught.
That single rule changes how unit economics should be read. A company with strong cash collection can still have weak revenue if delivery is slow and dropouts are high.
How does a refund window change CAC payback?
CAC, or customer acquisition cost, is the sales and marketing spend needed to win one paying learner. CAC payback is the number of months it takes for the gross margin from that learner to cover the CAC. A refund window sits directly in the path of that payback.
If a learner can claim a full refund within a stated period, then no revenue is truly earned until that window closes. Marketing spend, however, is incurred upfront. The practical consequence: the payback clock should start when the refund window shuts, not when the payment is collected.
Three habits make this worse:
- Counting enrolments at payment as "customers acquired" before the refund window closes.
- Spreading CAC across gross bookings rather than net revenue.
- Ignoring the cost of processing refunds, chargebacks and support tickets.
What is a cohort dropout curve and why does it matter?
A cohort is a group of learners who enrolled in the same period. A dropout curve tracks what share of that cohort is still active, and still paying, at each point in the course. In subscription or instalment models, the curve is the revenue curve.
Most edtech dropout curves are steep early and flatten later. Learners leave in the first few weeks, often before the first assessment. A company that assumes average retention across a 12-month course will overstate revenue in the early months, exactly when CAC needs to be recovered.
The fix is to build the curve from actual cohort data, month by month, and to apply it separately for each product and price point. A cheap self-paced course and an expensive mentored programme behave very differently.
A simple way to model net revenue per learner
- Start with the list price and subtract any discount or scholarship actually given.
- Apply the refund rate observed inside the refund window.
- Apply the cancellation rate for instalment or subscription plans.
- Apply the dropout curve to the remaining learners, month by month.
- Recognise revenue only for the months or modules actually delivered.
- Subtract variable delivery costs — teaching, mentoring, content, support — to get contribution margin.
- Compare cumulative contribution margin against CAC to find the true payback month.
Why do cancellation rates and refunds get confused?
A refund is money returned after a payment. A cancellation is a learner stopping a recurring plan or an instalment schedule. They overlap but are not the same. A learner may cancel without ever asking for a refund of past months, or seek a refund without formally cancelling.
For unit economics, both reduce net revenue and both should be tracked separately. Refunds usually cluster near the start; cancellations accumulate through the course. Mixing them into one "churn" number hides where the money leaks.
What should a founder check before quoting CAC payback?
Before putting a payback number in an investor deck, check the following:
- Is the payback calculated on net revenue after refunds, or on gross bookings?
- Does the refund window overlap the payback period?
- Is the dropout curve based on real cohorts or on an assumed average?
- Are delivery costs per active learner included in gross margin?
- Is deferred revenue treated as a liability, not as earned income?
- Are refunds and cancellations reported separately in management accounts?
What it means for Indian edtech operators
India's learning market spans test-prep, upskilling, K-12 supplements and language apps, and each carries different refund norms and dropout patterns. The common thread is that cash collected is not revenue earned. A company that reports enrolments and gross bookings without a dropout curve is, in effect, assuming every learner stays to the end.
Investors should ask for cohort-level net revenue, refund rates inside the stated window, and a payback calculation that starts after the window closes. Founders who build these numbers into their dashboards early will make better calls on pricing, refund policy and where to spend on acquisition. The metric that matters is net revenue per enrolled learner, tracked cohort by cohort, not the size of the booking number.
In this story
Frequently asked questions
What is edtech refunds and dropout economics?
It is the practice of measuring edtech revenue only after subtracting refunds inside the stated window, cancellations of instalment or subscription plans, and learners who drop out before completing the course. It gives a realistic view of CAC payback and cohort profitability.
Why is net revenue per learner better than gross bookings?
Gross bookings count every rupee sold, including money later refunded or never earned because a learner left. Net revenue per learner reflects what the company keeps after refunds and dropouts, and only for the portion of the course actually delivered.
When should CAC payback start in edtech?
It should start when the refund window closes, because until then the fee can be returned. Marketing spend is incurred at enrolment, so counting payback from the payment date flatters the number and hides risk.
How do you build a cohort dropout curve?
Track each enrolment batch month by month and record what share of learners is still active and paying. Build separate curves for each product and price point, then apply them to forecast revenue instead of using a single average retention rate.
Does revenue recognition affect reported profitability?
Yes. Fees collected in advance sit as deferred revenue until the course is delivered. Profits reported on cash collected rather than on delivered service overstate how much a cohort has actually earned.
