A refund policy written for a physical product or a simple software subscription does not translate cleanly to a course. A course is not returned the way a product is, and it does not stop being used the moment someone cancels the way a subscription does. Instead it is consumed gradually, in modules, over weeks, sometimes alongside a live cohort with its own start date and its own set of classmates. A student who watched three of twelve modules is in a genuinely different position than one who watched none, and a student asking for a refund the day before a cohort starts is in a different position than one asking two weeks after it wrapped. Generic refund language does not draw those distinctions, and course creators who copy one in end up settling disputes case by case instead of pointing to a policy that already answers the question.
Self-Paced Courses vs. Live Cohorts
The first decision is which format the refund rule needs to fit, because self-paced and cohort-based programs raise different failure modes. A self-paced course has no start date to anchor a cutoff against, so its refund window is usually measured from purchase or from how much of the content a student has accessed. A live cohort has a fixed start date, group sessions, and often a capped number of seats, so its refund window is usually measured against that start date instead, since a seat held for someone who never intends to show up has a real cost to the business (a spot another paying student could have taken).
Refund Windows by Program Format
| Self-paced course | Live cohort program | |
|---|---|---|
| Refund trigger | Time since purchase or content accessed | Time relative to cohort start date |
| Typical window | 7 to 14 days from purchase | Full refund until start, then prorated |
| Common threshold | Percent of modules or video watched | Which live session has already occurred |
| Main risk without a policy | Finishes course, then requests refund | Seat held, then abandoned before start |
If a business sells both formats, the policy should say so explicitly and state each rule under its own heading rather than writing one blended paragraph that leaves a reader guessing which sentence applies to which product. A student who bought a self-paced library and a student who bought a live cohort seat should each be able to find their own rule without cross-referencing the other section.
Partial-Completion Refund Thresholds
For a self-paced course, the most defensible refund rule ties eligibility to how much of the content a student has actually accessed, not just how many days have passed since purchase. A flat time-based window (say, 14 days) sounds simple, but it lets someone finish an entire course inside the window and still request a full refund, which is the scenario course creators run into most often once a policy has been live for a while.
A completion threshold fixes this by capping refund eligibility at a stated percentage of the course accessed, commonly somewhere between 20% and 30%. Below that threshold, a student who tried a couple of modules and decided the course was not a fit gets a refund. At or above it, the platform treats the content as substantially consumed and the purchase as final. This requires the platform to actually track module or video completion (most course platforms, including Teachable, Kajabi, and Thinkific, expose this data), and the policy should name the metric it uses, percentage of video watched, modules marked complete, or lessons opened, so there is no argument later about how “20%” gets measured.
Bundling the completion threshold with a time window (for example, “within 14 days of purchase and less than 20% of the course accessed”) is the more common real-world pattern, because either rule alone has a gap: a pure time window lets someone binge the whole thing before day 14, and a pure completion threshold lets someone sit on an unused purchase for a year and still ask for money back. Requiring both closes each gap.
Cohort-Start Cutoffs
For a live cohort, the cleanest structure ties refund eligibility to the start date rather than to a fixed number of days, since cohorts run on their own calendar and “14 days from purchase” means something different for someone who registers six months early versus six days before the cohort begins.
Many cohort programs also collect a deposit at registration to hold the seat, and it is worth stating up front whether that deposit is refundable at all, since some businesses treat it as non-refundable specifically to discourage seat-holding by students who are not committed. If the deposit is non-refundable, say so in the same clause that states the rest of the refund window, not buried in a separate payment-terms section a student is unlikely to read before registering.
A reasonable cutoff for a cohort in progress is the second live session: a refund request made before or during the first one or two sessions gets a partial refund (the business has delivered some value but the student has barely started), while a request made after that point does not, since the cohort has moved past the point where a new student could realistically catch up anyway. Where exactly to draw that line depends on cohort length. A four-week cohort and a twelve-week cohort should not use the same cutoff, and the policy should name a specific session number or week, not a vague “early in the program” that has to be interpreted after the fact.
The No-Refund-After-Access Threshold
Distinct from partial-completion tiers, most course refund policies also need a hard stop: a point past which no refund is available regardless of how little was watched, tied to whether the student downloaded gated materials. A course that includes downloadable worksheets, templates, swipe files, or a completed project file creates a different problem than streamed video, because once those files are downloaded, the platform cannot verify or reverse that they were accessed, and a refund at that point functions as a way to get the deliverables for free.
The clean way to write this is a standalone clause rather than folding it into the completion-percentage rule: “Refunds are not available once any downloadable course materials, templates, or completed project files have been accessed, regardless of overall course completion percentage.” This closes a gap that a pure video-completion metric misses entirely, since a student could watch 5% of the video content, download every template in module one, and still technically qualify for a refund under a rule that only tracks video-watched percentage.
Chargebacks and Platform-Specific Rules
If a course sells through a marketplace platform (Teachable, Kajabi, Thinkific, Podia, or a payment processor’s own dispute system) rather than a fully custom checkout, the platform’s own refund and chargeback rules sit on top of whatever the creator’s policy says, the same way an app store’s refund rules override a mobile app’s own terms. A student who disputes a charge with their card issuer directly, instead of requesting a refund through the platform, can trigger a chargeback that bypasses the policy’s stated window and threshold entirely, since the card network’s dispute process runs on its own timeline and its own criteria. Stating the refund policy clearly and requiring students to request refunds through it before disputing a charge does not prevent every chargeback, but it does give the business something concrete to submit as evidence if a dispute is contested, which is usually the difference between winning and losing a chargeback fight.
Putting the Policy Together
A complete refund policy for a course or cohort program states, in order: which format the rule applies to when a business sells both, the completion or time threshold for self-paced content, the cohort-start cutoff and deposit terms for live programs, the hard no-refund-after-access line for downloaded materials, and how disputes are meant to be raised before they escalate to a card chargeback. Each piece closes a specific gap the others leave open, and skipping any one of them is usually where a support inbox ends up litigating the same disagreement over and over instead of pointing to a line the student already agreed to.
Our refund policy generator builds this structure from a short set of questions about format, thresholds, and cohort dates, so the completion percentage and cutoff rules are stated as actual policy text rather than left implied. If the course also needs a full Terms and Conditions covering access windows and student conduct rather than just the refund clause, see Terms and Conditions for an Online Course or Cohort-Based Program, and if it is still unclear whether “refund” and “return” are even describing the same thing for a digital product, Refund Policy vs Return Policy covers the distinction.
