构建假日发现：Pinterest 工程师如何重新构想礼物发现

Megan Blake, Usha Amrutha Nookala, Jeremy Browning, Sarah Tao, AJ Oxendine, Siddarth Malreddy

Overview & Context

概述与背景

The holiday shopping season presents a unique challenge: helping millions of Pinners discover and save perfect gifts across a vast sea of possibilities. While Pinterest has always been a destination for gift inspiration, our data showed that users were facing two key friction points: discovery overwhelm and fragmented wishlists. With 85% of weekly US Pinners having made a purchase based on Pins from brands¹, we saw an opportunity to create a more streamlined gift discovery experience that meets Pinners where they already are — at the early stages of shopping inspiration.

假日购物季节带来了独特的挑战：帮助数百万的 Pinners 在广阔的可能性海洋中发现和保存完美的礼物。虽然 Pinterest 一直是礼物灵感的目的地，但我们的数据表明，用户面临两个主要的摩擦点：发现的压倒性和碎片化的愿望清单。由于 85% 的每周美国 Pinners 基于品牌的 Pins 进行了购买，我们看到了创造更流畅的礼物发现体验的机会，以满足 Pinners 在购物灵感早期阶段的需求。

The Challenge

挑战

Holiday gift shopping traditionally involves juggling multiple browser tabs, wishlists, and recommendations. Pinterest users actively look for new brands and ideas, with 96% of top searches being unbranded², showing Pinners are open to discovering fresh products. The engineering challenge was clear: how could we create a unified experience that maintains the serendipity of Pinterest discovery while adding structure to the gift shopping journey?

假日礼物购物传统上涉及同时处理多个浏览器标签、愿望清单和推荐。Pinterest用户积极寻找新品牌和创意，96%的热门搜索是无品牌的²，显示出Pinners愿意发现新产品。工程挑战显而易见：我们如何创造一个统一的体验，既保持Pinterest发现的偶然性，又为礼物购物旅程增添结构？

Our solution needed to:

我们的解决方案需要：

Leverage our existing personalization infrastructure while introducing gift-specific optimizations
利用我们现有的个性化基础设施，同时引入礼物特定的优化
Create a seamless saving experience that automatically generates shoppable wishlists
创建无缝的保存体验，自动生成可购物的愿望清单
Build a UI that feels fresh and distinct from our standard feed experience
构建一个感觉新鲜且与我们标准信息流体验不同的用户界面

Technical Foundation

技术基础

We approached this challenge by building on three core tec...