Cut optional fields on the pay step

🎯 Hypothesis

• If we strip optional fields from the pay step
• By keeping only email and card, framing Pay as keeping free shipping, and putting the refund on the same row
• Will more visitors complete checkout
• Because Choice Overload removes extra decisions, Loss Aversion makes the CTA about what they keep, and Risk Reversal lowers the last hesitation

🔬 Research

Psychology / behavioural economics

• Extra options slow the decision and raise abandonment; choice overload is the leak when company, coupon, and address sit in front of the card. - source
• Losses weigh about twice equivalent gains, so loss aversion belongs in the Pay label ("Keep free shipping") rather than a generic verb. - source

Case studies

• Large checkout studies still find more fields on the pay step than people will complete; cutting optional inputs is the reliable move. - source
• GOV.UK service patterns put one job on the page so people can finish without ranking extras. - source

🖼️ Control

The existing UI for the pay step.

✨ Treatment

Removes company, billing address, and coupon from the pay step. Pay copy becomes "Keep free shipping · Pay $84.00", with a 30-day money-back line under the button.

<section class="pay">
  <label>Email <input type="email" value=""></label>
  <label>Card <input type="text" inputmode="numeric" value=""></label>
  <button type="button">Keep free shipping · Pay $84.00</button>
  <p>30-day money-back. Cancel anytime.</p>
</section>

.pay { max-width: 22rem; padding: 1.5rem; font: 16px/1.4 system-ui, sans-serif; }
.pay label { display: grid; gap: 0.35rem; margin: 0 0 0.85rem; font-size: 13px; color: #444; }
.pay input { height: 2.5rem; padding: 0 0.7rem; border: 1px solid #ccc; border-radius: 8px; }
.pay button { width: 100%; height: 2.75rem; border: 0; border-radius: 8px; background: #111; color: #fff; font-weight: 600; }
.pay p { margin: 0.75rem 0 0; font-size: 13px; color: #555; }

📊 Metrics

Primary metric

The deciding metric.

• Checkout completion (%) — share of pay-step visitors who successfully pay

Measurement metrics

Help explain the result.

• Time to pay (avg) — seconds from pay-step view to successful payment
• Field error rate (%) — share of pay-step visitors who see a validation error

Guardrail metrics

Must not regress during the experiment.

• Support contacts about checkout (count) — tickets tagged checkout or payment
• Refund rate (%) — share of completed orders refunded within 7 days

📈 Results

✅ Conclusion

Variant shipped to all users? Control / treatment