Refunds and the regulatory reality

#1. What's actually verifiable on refund rates, and what isn't

Searching for course-industry refund rates surfaces a wide, mutually contradictory range — "under 2%," "under 5%," "5–10%," "up to 21–30%," "18–28% and rising," "chargebacks up 41% year over year." Nearly every specific figure at the higher end of that range traces to content-marketing pages for refund-prevention or student-retention tooling — sites with a direct commercial interest in the problem looking as large and urgent as possible — and none disclosed a sample, a data source, or a methodology this research could independently check. This course does not repeat a specific industry-wide refund-rate percentage as fact, for the same reason it declined to repeat a specific completion percentage in the previous lesson: no source cleared the bar.

The individual creator disclosures at the low end of that range (a handful of course sellers publicly reporting "under 2%" or "under 5%" refund rates on their own blogs) are more credible in the narrow sense that they're specific, named, and dated — but they're self-reported, unaudited, single-business figures from people with an obvious incentive to publicize their best number, not a representative industry sample. [Speculative] across the entire refund-rate question — this is a genuine, stated gap in this research, not a number smoothed over.

#2. Chargebacks: a real, weakly-sourced pattern worth naming honestly

Multiple independent payment-processor blogs (covering merchant chargeback-rate benchmarks by industry) report the "education and training" merchant category at roughly 1.02% chargebacks, meaningfully above the roughly 0.60% all-industry average they each separately cite. [Directional], and named as a specifically weak Directional rather than treated as solid: the pattern recurs across several independently-published sources, which is more than a single unverified claim earns — but not one of those sources cited an original data provider, a sample size, or a methodology behind the number, so what's being corroborated is the retelling, not necessarily the underlying data. The direction is plausible on its own mechanistic terms regardless of the exact figure: education-and-coaching purchases tend to be higher-ticket, delivered with a delay (access to a course, not an instant physical good), and vulnerable to the specific disputed-value pattern chargeback researchers associate with intangible services generally — a buyer who's dissatisfied with an intangible outcome has an easier case to make to their card issuer than a buyer disputing a tracked physical shipment.

Practically: staying meaningfully under the card networks' own chargeback-rate thresholds (roughly 0.9% for Visa, higher for other networks, above which a merchant faces monitoring programs and eventually processing restrictions) is a real operational constraint for a course business, not a hypothetical one, given that this category's reported rate sits close to those thresholds even in the most charitable reading of a weakly-sourced number.

#3. The regulatory backdrop, 2026: two live threads

Earnings claims. The FTC's January 2025 proposal to expand its Business Opportunity Rule to cover "business coaching opportunities" — any program represented as teaching someone to build or operate a business — would, if finalized, require written substantiation for any earnings claim, available to a consumer on request. [Established] as to the proposal existing and its text (ftc.gov NPRM, January 13, 2025); as of this research it remained proposed, not finalized — a real signal of regulatory direction, not yet a binding rule for every course seller. It sits directly downstream of the FTC's already-completed enforcement in this exact space: the previous lesson covers the MOBE and Digital Altitude cases in detail, both of which were prosecuted under the FTC's existing, general unfair-and-deceptive-practices authority (FTC Act Section 5) rather than a course-specific rule — meaning a course seller making unsubstantiated income claims is already exposed today, proposed rule or not.

Cancellation and negative-option billing. A course sold as a recurring membership (the Circle/Skool model from module 2) is a negative-option plan under FTC terminology — a subscription that renews and bills automatically unless the customer affirmatively cancels — and this is a live, actively-moving regulatory area in 2026. The FTC's "click-to-cancel" rule, which would have required cancellation to be at least as easy as signup, was vacated by the Eighth Circuit on July 8, 2025 on procedural grounds (the FTC hadn't conducted a legally-required preliminary regulatory analysis), and is no longer in effect. [Established] — multiple independent law-firm regulatory trackers report the same vacatur date and procedural basis, consistent with each other. The FTC submitted a draft Advance Notice of Proposed Rulemaking to revive a version of the rule on January 30, 2026, with a public comment period running to April 13, 2026. [Established] as to the filing and comment-period dates; the eventual shape of any revived rule is genuinely unresolved as of this research. The underlying general prohibition on deceptive or unfair negative-option billing practices (under the FTC's older Restore Online Shoppers' Confidence Act authority, which the rule vacatur did not touch) remains active enforcement law regardless of the specific rule's status — a course-membership seller building an easy-signup, hard-to-cancel funnel is exposed under existing law today, independent of whether the click-to-cancel rule specifically is ever finalized.

#4. What this means practically

Two defensible, low-regret positions follow from a regulatory picture this unsettled: keep cancellation at least as easy as signup regardless of what the rule currently requires, because the underlying deceptive-practices exposure predates and outlives this specific rule's on-again-off-again status; and keep genuine, checkable substantiation for any income or outcome claim in your own marketing — real, dated, named examples with your actual numbers, not aggregated or unaudited figures — because that's the exact standard the proposed Business Opportunity Rule expansion would impose if finalized, and it's good practice under existing Section 5 authority regardless.