How A/B Testing Transforms JD Express Mini‑Program Design: From Basics to Real‑World Results

Design as a Science: Using A/B Testing in JD Express Mini‑Program

1. Basics

A/B Testing Definition

A/B testing (also called split testing or bucket testing) compares two versions of a webpage or app to determine which performs better. Users are randomly divided into a control group and a treatment group, and the core metrics of each group are compared to evaluate the benefit of a design change.

Why Conduct A/B Testing

Use test data to assist decision‑making and identify problem focus.

Build hypotheses to understand why element changes affect user behavior, capturing user preferences to guide future design.

Through long‑term experiments, explore variables that drive data growth (style, layout, decorative elements, benefit exposure) and create reusable design specifications.

Designers often encounter UI style A/B tests, which are the focus of this article.

How to Conduct A/B Testing

Define Experiment Goals – set clear metrics such as click‑through rate of a resource slot, coupon claim volume, or activity share count.

Create Experiment Variables – modify visual elements (button color, copy, amount size, visual style, main object placement) or functional elements (show/hide features, change operation steps).

Generate Hypotheses – decide which variables will be core indicators and which will be auxiliary.

Design Experiment Layout

Single‑factor design: only one variable changes (e.g., color or copy).

Multi‑factor design: multiple variables change, producing combinations (e.g., image style A/B × benefit highlight A/B).

Collect Data and Allocate Traffic – randomize users to groups, ensure the experiment does not harm user experience, and adjust traffic allocation based on experiment risk and expected ROI.

Analyze Results – compare control and test groups; if key metrics improve, consider full rollout or further refinement; if not, iterate.

2. Exploration: JD Express Mini‑Program Order‑Completion Page "Lucky Red‑Envelope" Popup Test

The popup promotes a red‑envelope activity, offering coupons that can be shared for extra rewards. The old version suffered from outdated visual style and unclear information hierarchy.

Optimization goals included a more prominent visual presentation, stronger button cues, clearer benefit hierarchy, and enhanced atmospheric effects.

Adjust visual form to make the main element stand out.

Enlarge buttons and highlight the share button.

Emphasize benefit points and increase their visual weight.

Use lighting effects to create an immersive atmosphere.

Data comparison of 10 days before launch and 20 days after showed a significant increase in popup click‑through rate and improvements in share rate and coupon claim rate.

Why Test the Order‑Completion Popup?

Users who reach the order‑completion page have already placed an order, providing a clean behavior set for measuring subsequent actions.

3. Practical: JD Express Mini‑Program Home‑Page Popup A/B Tests

Project Background

The home‑page popup is the most prominent entry point, receiving the highest traffic. It is used for various promotions such as holiday coupons, weekly coupons, weekend coupons, and family‑card offers.

First Round

Key metrics: click‑through rate of the resource slot and coupon claim volume.

Experiment variables focused on whether the benefit point is highlighted, the amount is enlarged, and the coupon display style is more attractive.

Second Round

Explored the impact of coupon quantity and the presentation of discount versus amount.

Multiple‑coupon (grid) vs. single‑coupon (large) display.

Benefit point shown as discount vs. monetary amount.

4. Conclusions and Future Planning

Visual style updates increase user freshness and click‑through rates.

Highlighting benefit points (especially discounts) drives higher engagement than merely showcasing the activity theme.

More coupons displayed lead to higher click‑through rates.

When visual style is constant, discount presentation outperforms amount presentation.

Continuous iteration and monitoring are required because user responses evolve with visual updates, and the optimal solution is product‑specific.