01 / The issue
Why this decision deserves attention.
Tiller offers a paid annual subscription after a free trial. When new trials declined, the team tested copy and rebuilt the payment step.
Instead of continuing to maintain its own embedded card form, Tiller adopted Stripe Checkout, adding familiar wallet and password-manager flows.
Stripe reports a 90% increase in trial sign-ups, a three-times faster path from first login to trial and more than 1,000 lines of code removed. These are Stripe’s customer-story results, not a general guarantee.
02 / The principle
What the evidence tells us.
A custom checkout can feel like product polish while quietly adding trust, validation and maintenance problems. Early products usually benefit from a proven hosted flow until there is evidence that custom UI is the constraint.
03 / The decision
What to change in the way you work.
- 01
Reduce the number of unfamiliar decisions at the payment moment.
- 02
Use a maintained checkout for wallets, validation and payment-method updates.
- 03
Measure the full path from pricing-page intent to successful payment—not button clicks alone.
04 / How Ten Million helps
Pricing & checkout
For a product at this stage, we turn the lesson into concrete working material:
- Pricing model and plan-structure review
- Stripe Checkout or Merchant of Record architecture
- Trial, coupon, tax, cancellation and failed-payment flow map
- Test-mode checklist and conversion-event instrumentation
05 / Builder playbook
What to do this week.
- 01
Choose the simplest payment architecture that covers the countries and tax model you need now.
- 02
Test success, failure, retry, cancellation and webhook replay paths.
- 03
Show renewal terms and cancellation expectations before payment.
- 04
Instrument each step so a checkout problem is distinguishable from a pricing problem.