# Greeting card pricing and add-on flow changes: +24% paid card sales

- Outcome: Won
- Page type: Checkout (Ecommerce)
- Elements: Upsells & cross-sells, Pricing, discounts & savings display
- Test run by: [Optimizely](https://liftarchive.com/practitioners/optimizely/)
- Commentary by: LiftArchive (our analysis, not theirs)
- Original source: https://www.optimizely.com/field-notes/articles/experimentation-use-cases-retail-checkout-pages
- Page: https://liftarchive.com/test/optimizely-greeting-cards-checkout-bandit/

Changing greeting card pricing and the checkout add-on flow lifted paid card purchases by 24% and average order value by 8.4%. Optimizely ran this test for an online greeting card retailer and published it as an Optimizely Field Notes use case. Several changes were tested together using a multi-armed bandit, so the result can't be credited to any single change.

## Result

**+24% paid card sales**

- +24% paid card purchases. More shoppers paid for a card rather than skipping it.
- +8.4% average order value. Orders got bigger on average.
- Both lifts come from a bundle of pricing and flow changes, and the winning details are only loosely described.

## Control

The existing checkout flow and the current greeting card pricing.

## Variant

Several variations of greeting card pricing and the checkout flow. A multi-armed bandit sent more traffic to the best performers over time.

## Why it likely won (our reading)

- Pricing changes may have made a paid card feel like an easy add-on at checkout.
- A smoother add-on step may have put the card offer in front of shoppers at the right moment.
- A bandit moves traffic toward stronger variants as the test runs, which captures gains sooner.

## Caveats

Bundled changes across several variants in a multi-armed bandit, so no single element can take credit. Vendor-reported, with no sample size, duration or confidence level shared.

## How to run this test

- Hypothesis: If we adjust card pricing and offer the card as a simple checkout add-on, then more shoppers will buy a paid card because the offer is cheaper to accept and easy to add.
- Primary metric: Paid add-on purchases per order
- Guardrail: Checkout completion rate
- Guardrail: Average order value
- Guardrail: Margin per order
- Who should test it: Gift and occasion stores with a natural add-on, such as cards, gift wrap or notes. A bandit suits short-lived offers where earning more during the test matters more than a clean read on each change.

---

Credit: Optimizely ran this test and owns the result. Summary and commentary by LiftArchive (https://liftarchive.com/test/optimizely-greeting-cards-checkout-bandit/).
Usage: Free to use, quote and build on, with credit and a link: name the practitioner or organization that ran the test and link to its LiftArchive page. The results belong to the practitioners who ran and shared them; the summaries and commentary are LiftArchive's.