Paginated Loading of Shopping Cart Auxiliary Information in JD E‑commerce Platform

The JD shopping cart has continuously expanded its capacity from 80 items in 2015 to over 220 items for premium users, causing significant backend load due to full‑data refreshes and complex calculations for selections, promotions, and pricing.

To avoid the performance bottleneck of loading all auxiliary information (such as coupons, stock, activity tags) at once, the article proposes a paginated loading approach that fetches auxiliary data only for the items currently visible on the screen and pre‑loads adjacent pages based on user scrolling behavior.

Three pagination strategies are evaluated: pure product pagination, store‑based pagination, and a combined product‑plus‑store pagination that keeps items from the same store together while still paging by product count, ultimately chosen for preserving business logic and aligning with upstream batch interfaces.

The solution includes a configurable pre‑loading policy that balances user experience against unnecessary API calls, a timestamp‑based dirty‑data check to discard stale responses, and adaptive loading based on scroll speed to further reduce unnecessary traffic.

Measured results show a reduction of over 30% in auxiliary‑information API calls, substantial resource savings during peak traffic, and a smoother, low‑latency shopping cart experience without sacrificing functionality.