skills/offers/references/guarantee-design.md
A guarantee directly raises perceived likelihood of achievement (the buyer thinks: "they'll only offer this if they're confident") and lowers effort & sacrifice (less emotional risk). It's one of the highest-leverage levers in offer design.
The wrong guarantee hurts more than no guarantee. Pick the type that matches your business model.
| Type | What it promises | When it works | When it backfires |
|---|---|---|---|
| 1. Unconditional money-back | "Refund anytime within X days, no questions" | Low-priced info, high-trust audience | High-priced/high-touch; refund risk eats margin |
| 2. Conditional money-back | "Refund if you complete X and still don't see Y" | Courses, programs requiring effort | Sophisticated buyers; harder to honor publicly |
| 3. Better-than-money-back | "If it doesn't work, full refund + $X" | Confident delivery, ample margin | If you fail; the few failures explode publicly |
| 4. Service-level / SLA | "If we don't deliver X by Y, your money back" | Productized services, agency work | Vague SLAs you can't measure |
| 5. Performance-based | "Pay only when X happens" (rev share, results-based) | Sophisticated B2B, high-confidence delivery | Long cycles, hard-to-attribute outcomes |
| 6. Anti-guarantee | "No refunds. Make sure you want it." | Premium audiences, mature buyers | Confused / first-time buyers; reads cold |
| 7. Outcome-or-extension | "If you don't get X by Y, we continue free" | Coaching, services with extendable time | Open-ended cost; choose with care |
| 8. Comparison guarantee | "Beat [competitor]'s result or refund" | When you can credibly compare | When you can't measure the competitor cleanly |
Decision tree:
What's your buyer's biggest perceived risk?
What's your refund tolerance?
What's your buyer sophistication?
How measurable is the outcome?
Strong: "Complete all six modules within 60 days, submit the final exercise, and if you haven't [specific outcome] we refund in full." Conditional on effort, clear on outcome.
Weak: "100% money-back guarantee." No conditions = refund magnet for buyers who never engaged.
Strong: "After the first two sessions, if you don't think the engagement will deliver, we end it and refund the unused balance." Mid-engagement off-ramp builds trust.
Weak: "Satisfaction guaranteed." Means nothing.
Strong: "First month is a paid pilot. At the end, if you don't see [specific milestone], you don't pay for month 2 and we end on good terms." Clear gate, clear out.
Weak: "We'll work until you're happy." Open-ended cost. Don't.
Strong (premium audience): "No refunds. The application process is rigorous because the value is real. If you're not sure, don't apply yet." Anti-guarantee works here.
Weak (premium audience): Generic 30-day refund. Reads cheap.
Strong: "30-day no-questions refund." The transactional bar is "I bought it, looked at it, didn't want it." Unconditional fits.
Weak: No guarantee. The buyer's risk is too high for the price.
Strong: Free trial or annual-prepay-with-money-back-in-first-30-days. Reduces friction without locking in unhappy users.
Weak: "Cancel anytime" alone — not a guarantee, just standard SaaS terms.
Strong: Double-your-money-back or comparable risk inversion. Direct-response buyers expect risk-reversal-heavy offers.
Weak: Vanilla 30-day refund. Doesn't differentiate from every other ad on the platform.
The guarantee text matters. Patterns that work:
Specific terms:
If, after completing the first 4 weeks of the program, you can't point to one specific business outcome you've achieved, email us and we'll refund 100%.
Confident tone:
We know this works. If it doesn't for you, we don't want your money.
Acknowledge the awkwardness:
Guarantees feel slimy. Here's ours anyway: if you do the work in modules 1–3 and don't see meaningful traction, we refund.
Patterns that don't work:
"Double your revenue or your money back + $1,000." If even 1 in 50 buyers fails and gets the bonus refund + writes a public review, the offer is permanently damaged.
Stress-test: what happens if 10% of buyers invoke the guarantee?
"Refund if you watched all 24 modules, completed the 6 exercises, attended every live call, and posted in the community at least once per week."
Buyers read this as "they made it impossible to actually get a refund." Trust drops.
Two conditions max. Three only if they're closely related (e.g., "completed the course AND submitted the final project AND emailed us a question").
If your guarantee is your strongest perceived-likelihood lever, put it on the sales page in 24pt text. Move it above the buy button.
Re-read your guarantee text every six months. The wording that worked a year ago may now be undermined by something you've changed about your offer.
Strong proof + weak guarantee > strong guarantee + weak proof. Order matters. Build proof first, then layer on the guarantee.
Anti-guarantees ("no refunds, this is final") work when:
What you're saying: "We don't need a guarantee because the work is real, the buyer has self-qualified, and we won't engage in transactional refund games."
The wrong audience reads this as cold or scammy. The right audience reads it as confidence. Know your buyer.
When auditing an offer with no guarantee (or a weak one), ask:
Most offers don't have the wrong guarantee — they have no guarantee at all. Adding any guarantee is almost always a lift. Adding the right one is the lever.