Measure the effect with a controlled test
Establish a baseline before changing the checkout. Useful measures include the share of sessions that reach payment selection, payment attempts, successful orders, error rates, method selection, and mobile versus desktop completion. Keep the checkout design, traffic sources, and promotional conditions as stable as practical while testing one meaningful change.
If the checkout platform supports an experiment, compare a control group with a group that sees the added method. Stripe documents payment-method A/B testing for eligible integrations and recommends measuring conversion, average order value, and shifts among methods. For other systems, use the provider's supported testing approach. Allow enough traffic for a useful comparison, watch technical errors as well as orders, and avoid declaring success from a short spike.
Review the full effect. A method that attracts selections but creates more failed attempts or customer-service work may not improve the overall checkout. A method used by a smaller segment can still be worthwhile when it serves an important audience without making the main flow harder.