Pre-checked subscription: -9.17% RPV, subscribers 4.7% to 65.8%
Pre-checking the subscription option on the product page cut first-order revenue per visitor by 9.17%, while the share of orders that were subscriptions jumped from 4.7% to 65.8%. Dave Diederen ran this test. Whether it was a win depends on repeat orders and retention, which the post does not report.
- Revenue per visitor: -9.17% ($3.41 to $3.09), range -15% to -3%. A clear first-order loss.
- Profit per visitor: -11.92% ($2.34 to $2.06). Upfront profit fell more than revenue.
- Conversion rate: -2.86% (5.26% to 5.11%), range -7% to +1%. A small drop that may partly be noise.
- Average order value: -6.50% ($64.79 to $60.58). Subscribers bought fewer units per order (3.07 to 2.66).
- Subscription share of orders: 4.7% to 65.8%, and subscription revenue per visitor rose from $0.16 to $1.97. Most buyers stayed on the default.
The change
The variant pre-selects the subscription option instead of one-time purchase.
Simplified mockup drawn by LiftArchive from the shared screenshots.
Control
The product page with one-time purchase as the default. Only 4.7% of orders were subscriptions, and the post does not show how the option was presented.
Variant
The subscription option was selected by default, so shoppers had to switch to one-time purchase themselves. The page design was not shown.
Why it likely lost
Our reading of the result, not the practitioner's claim.
- a pre-checked subscription asks for a bigger commitment, so some one-time buyers hesitated and conversion dipped.
- shoppers who know another shipment is coming have less reason to stock up, which fits the fall in units per order.
- defaults are powerful. Most buyers kept the pre-selected option, which is why the subscriber share rose so sharply.
Caveats: This shows first-order results only; rebills, churn and lifetime value were not shared, so the long-term outcome is unknown. Sample size and test length were also not shared.
Who ran it
Dave Diederen (@DaveDiederen) shared this test on X. The numbers and screenshots are theirs.
View the original post on X
How to run this test on your store
- Hypothesis
- If we pre-check the subscription option, then far more buyers will subscribe because most people keep the default, trading some first-order revenue for repeat orders.
- Primary metric
- Profit per visitor over 60 to 90 days, including rebills
- Guardrails
- First-order revenue per visitor
- Subscriber churn and cancellation rate
- Refunds and chargebacks
- Who should test it
- Stores selling consumables people reorder, such as supplements, coffee or skincare, that can track rebills and lifetime value. It probably won't pay off for one-off purchases, or for stores judging tests only on first-order revenue.
Questions
Should the subscription option be pre-selected?
It sharply raises sign-ups: here subscriptions went from 4.7% to 65.8% of orders. But first-order revenue per visitor fell 9.17%, so it only pays if subscribers keep reordering.
Why did average order value drop with subscriptions?
Subscribers bought fewer units per order (3.07 down to 2.66). People who expect another delivery have less reason to buy several at once.
How should you judge a subscription A/B test?
Dave Diederen argues for comparing the upfront profit lost against the recurring profit gained. That means tracking rebills, churn and 30, 60 and 90-day lifetime value, not just day-one revenue.
Similar tests
@oliverkenyon@oliverkenyon
@DaveDiederen@DaveDiederen@DaveDiederen@DaveDiederen