Checkout Delivery-Option Selector — A/B testing pickup-point defaults

A checkout delivery step that pre-selects the nearest pickup point instead of home delivery, used by an online shopper completing an order, with the single job of measuring whether the default changes what they pick.

The customer — landing on the delivery step of checkout after adding items to the basket. Sessions are randomised into one of two buckets on arrival; a product manager can also force a bucket from the demo screen to inspect each experience.

1. Screen 1: Bucket assignment — Control or Variant
2. Screen 2: Delivery step with a pre-selected option and basket summary
3. Screen 3: Pickup-point list (distance, ready-by time, opening hours, saving)
4. Screen 4: Order confirmation — records the completed session
5. Screen 5: Results dashboard comparing the two buckets

• Control: home delivery pre-selected, pickup points listed below with no incentive copy.
• Variant: nearest pickup point pre-selected plus "Save €1.50 · Ready tomorrow" — default bias with an explicit reason.
• Home delivery stays one click away in every bucket. The test measures default behaviour, never restriction of choice.
• Nearest point = smallest distance from the mocked geolocation. If no point is within 2 km, all buckets fall back to home delivery.
• Checkout completion rate is a hard guardrail: a cost win that drops completion below the control band is a failed test, not a win.

All sample/mock data, stated explicitly: a mocked geolocation (Nørrebro, Copenhagen), three nearby pickup points with distance, ready-by time, opening hours and saving; one saved home address with an ETA and a €3.50 delivery fee; a fixed €74.90 basket. Session results are stored in the browser only.

• Empty — no sessions recorded yet on the results page
• Loading — placing the order (simulated latency)
• Error — pickup-point lookup fails, checkout falls back to home delivery
• Abandon — the session is recorded as not completed
• Success — order confirmed and the result written to the results store

Matches this portfolio: light, high-contrast surfaces, rounded cards on a neutral palette, restrained accent colour, no illustration-heavy checkout chrome.

No auth, no payment processing, no real carrier or geolocation APIs, no backend persistence, no basket editing or address book management, no statistical significance engine — the results page reports observed rates only.

Default-to-pickup increases pickup-point adoption and lowers average delivery cost per order, while checkout completion holds within guardrail. Primary metric: delivery cost per order. Secondary: pickup-point adoption rate and time-to-checkout-complete. Randomised at session/customer level.