The Platform

A B2B wholesale marketplace where buyers bid on auctions and pay after winning. Payment has always been buyer-initiated — meaning buyers manually log in and pay after an order is created. This creates a structural delay baked into every transaction.

Business Goal

Reduce order-to-pay time from ~2.2 days → under 24 hours to improve cash flow and fulfillment speed.

As long as payment is buyer-initiated, time-to-pay won't compress meaningfully. The only lever: make payment happen automatically after a bid is won.

Payment Landscape

Four payment methods in different stages — each with different speeds, risks, and adoption rates:

Initial Requirements

Direction set by Product & Leadership:

• Mandatory auto-charge once an order is created

• Start with one seller for a controlled pilot rollout

• Net Terms only to reduce risk/complexity and ship the MVP sooner

Research Focus

Before designing anything final, I identified three core unknowns to validate:

Existing User Flow

Payment was the final manual step — buyers had to actively return and pay after winning an auction.

Proposed Flow for Auto-Charge

Two changes to enable auto-charge:

• 1

  Mention auto-charge before bidding (at the Listing Detail Page) to set expectation

• 2

  Surface shipping & payment settings in the Bid Confirmation, so everything is set before the auction ends

Key UI — Entry Point

Setting expectation before bidding — an indication of auto-charge shown on the Listing Detail Page.

Entry Point — Listing Detail Page

Bid Confirmation — Shipping + Payment

A Constraint Surfaced a Better Question

During Stage 1 research, a critical finding emerged: wire transfer accounts for ~40% of GMV. These buyers couldn't be auto-charged (offline workflow), so the mandatory direction required exemption logic — thresholds, whitelists, or other workarounds.

This prompted a reframe: instead of exemptions for wire buyers, what if we gave all buyers a choice?

The Two Directions

Prototype Designs

Both directions were prototyped for A/B testing. Key screens: entry point on the Listing Detail Page and the Bid Confirmation modal.

Direction 1 — Mandatory

Direction 2 — Optional

A/B Test Setup

Participants were split into two groups to control for order bias. Both groups rated each direction and selected a preference.

Direction 2 Was Strongly Preferred

Opt-in Rate Still High With Choice

Even when given the option to opt out, buyers' expected opt-in rate was ~93.75% — meaning choice didn't sacrifice adoption.

Behavioral Observations

Launch — Learn — Iterate

Rather than betting everything on one direction upfront, I proposed a phased rollout to build evidence before committing.

Where It Landed

The research supported Optional. The shipped rule: ACH / Net Terms = auto-charge; Wire = manual. Rather than building complex exception logic for wire buyers, the team moved to a cleaner model — payment method itself becomes the opt-in mechanism.

Impact: shifted the team away from exception-heavy enforcement logic toward a more transparent, buyer-respectful design.

In Retrospective

In hindsight, I wish I'd run a quick "Optional" test even when the original scope was tightened to mandatory-only. Validating the alternative earlier might have changed the initial direction sooner.