Whatnot Hits $8 Billion GMV as Live Commerce Expands to 100 Categories

Whatnot GMV达80亿美元，直播电商拓展至100个品类 直播电商将购物转变为社交娱乐从而获得成功，实时的人际互动推动了冲动发现，并为小微企业带来了庞大销售额。 Grant LaFontaine 详细介绍了 Whatnot 的发展历程，从其童年收集卡牌的经历与早期域名谈判，到打造一个规模超80亿美元的直播电商平台。讨论对比了传统目的驱动型搜索购物与 Whatnot 基于发现的模式，在后者模式下，用户平均每天在直播互动中花费95分钟。LaFontaine 还概述了该平台在包括时尚、高尔夫和生鲜海鲜在内的100多个品类上的供给端扩张，并探讨了国际化增长、AI卖家工具以及稳健的实时信任与安全基础设施。 Grant LaFontaine 在宝可梦卡等收藏品买卖方面的背景，结合其在 YouTube 和 Facebook Marketplace 的产品经验，为 Whatnot 奠定了基础。 (0:00) Whatnot 始于2019年，旨在利用移动优先的社交电商来革新过时的收藏品市场，并以0.5枚 BTC 购得了该域名。 (3:49) 公司通过观察卖家设法利用社交媒体平台进行简陋的视频销售（且伴随脱节的支付和物流体验），从而捕捉到了直播电商的商机。 (9:18) 与 Amazon 或 eBay 等目的驱动型电商平台不同，Whatnot 的运作模式更像是一家专注于发现、娱乐和社区的数字购物中心。 (12:41) Whatnot 上超过80%的活跃用户在某一天内并不会购买商品，登录主要是为了社交互动，用户每天在该应用上的平均使用时长达95分钟。 (14:49)

Whatnot co-founder Grant LaFontaine argues that live shopping outperforms traditional e-commerce by replacing intent-based search with community-driven entertainment, expanding the total consumer market rather than merely capturing existing transactions. Through interactive streaming, the platform seeks to close the penetration gap between Asian live commerce and Western retail by rebuilding merchant trust across diverse merchandise categories. Entertainment-First Discovery Versus Intent-Based Search (12:41) LaFontaine distinguishes Whatnot's discovery-based model from traditional intent-driven e-commerce, where consumers search for specific SKUs. Drawing an analogy to window-shopping at a retail mall, Whatnot focuses on real-time entertainment, live bidding, and community interaction. According to platform data, over 80% of active users do not make a purchase on any given day, instead tuning in primarily to converse and watch streams, with users averaging roughly 95 minutes per day on the app. LaFontaine contends that this dynamic is demand-expansionary. While traditional search captures premeditated demand, live commerce prompts discovery and impulse purchasing across a wider audience. By prioritizing user enjoyment over rigid transactional efficiency, the platform shifts the core digital shopping experience from search bars to live interactive programming. Bridging the Adoption Gap and Category Growth (14:49) While live commerce constitutes 30% to 40% of retail e-commerce in China—representing roughly a $1 trillion market—it accounts for only single digits of the $1.2 trillion US e-commerce industry. LaFontaine asserts that this disparity is poised to narrow because live streaming offers merchants an efficient alternative to expensive physical retail leases, allowing zero-cost setups to reach global audiences immediately. Whatnot has diversified well beyond its initial niche of collectibles and Funko Pops into more than 100 product categories across 10 countries, including Canada, Japan, Australia, and five European nations. Women’s fashion has become the platform’s largest category, operating similarly to off-price retailers like TJ Maxx by helping sellers liquidate excess inventory at steep discounts. The company is now pursuing further expansion into high-value and regulated sectors, including vehicle walk-around auctions and alcohol. Merchant Economics and Live-Stream Trust Building (21:18) Traditional e-commerce platforms often reduce independent merchants to anonymous fulfillment channels, but live streaming foregrounds seller identity and personality. Whatnot processed over $8 billion in merchandise volume last year and surpassed that threshold again year-to-date, with top merchant operations scaling from solo teams to over 100 employees, generating up to $100 million in annual revenue at reported EBITDA margins of 20% to 40%+. The format also resolves fundamental consumer trust hurdles for difficult online categories. For example, a San Diego seafood distributor reached thousands of dollars an hour in sales by streaming directly from the docks. Live video allows buyers to inspect fresh catches, ask questions about handling, and verify origin in real time, overcoming the quality hesitation inherent to static product listings for perishable goods. Scaling Governance and Backend AI Integration To manage a user base scaling into tens of millions, Whatnot allocates approximately 40% of its total workforce to Trust & Safety. The platform deploys an automated, sub-second rules engine powered by large language models to analyze billions of live data points, flagging policy infractions such as harassment, shipping delays, or abnormal refund rates, and triggering automated sanctions based on seller history. Despite heavy investments in artificial intelligence, Whatnot deliberately avoids front-facing synthetic tools like AI avatars, which LaFontaine notes undermine the authentic human relationships essential to community retention. Instead, the company applies AI strictly to backend seller workflows, automating tedious cataloging tasks, inferring listing metadata, and delivering inventory performance analytics. Although Whatnot has demonstrated significant volume growth across niche and off-price goods, its long-term expansion into complex, heavily regulated sectors like automotive and alcohol will test whether live video can consistently manage complex logistics, legal compliance, and customer protection at global scale.

Whatnot 联合创始人 Grant LaFontaine 认为，直播电商通过将传统的搜索式购物转变为以互动娱乐和社区为核心的“发现式”体验，正在打破传统电商将卖家匿名化、工具化的弊端。凭借实时信任机制和极高的用户粘性，该平台正从最初的收藏品垂直赛道向全品类商业网络及全球市场快速拓展。 发现式娱乐驱动需求与用户高黏性 (12:41) 与传统电商高度依赖明确购买意图（如在搜索框输入特定鞋款）不同，Whatnot 的运作模式更接近线下商场漫步，用户主要出于消遣、寻宝和社交需求进入直播间。LaFontaine 强调，消费者并不在乎行业术语或市场规模，他们只关心过程是否有趣。在 Whatnot 上，超过 80% 的日活跃用户在当天完全不产生购买行为，而是专注于观看、实时聊天与社区互动，用户平均每日使用时长高达约 95 分钟，其参与度已可媲美主流社交媒体平台。 这种模式在本质上是对消费需求的扩张。在中国等亚洲市场，直播电商已占据零售总额的 30% 至 40%（规模约 1 万亿美元），而体量约 1.2 万亿美元的美国电商市场中，直播电商渗透率仍处在较低的个位数。LaFontaine 认为，直播电商凭借零边际开店成本和无限制的流量触达，必然会跨越这一市场鸿沟；对比仍拥有约 800 亿美元交易规模的 eBay，Whatnot 的直播形态能吸纳更多不愿入驻传统拍卖网站的成熟品牌与小型企业。 卖家核心地位与多元品类的爆发 (21:18) 传统电商平台在过去数十年间为了追求履约效率，往往隐去商家个体特征，将卖家降格为匿名的发货渠道。Whatnot 则反其道而行，将创业者的个人特质、品牌叙事和社群关系置于中心。平台去年商品交易总额（GMV）突破 80 亿美元，今年也已再度超过 80 亿美元大关。依托相对温和的抽佣政策，平台头部卖家的经营规模从一两人的微型团队扩展到百人以上，部分年收入可达上亿美元，息税折旧摊销前利润率（EBITDA margin）维持在 20% 至 40% 以上。 在业务品类上，Whatnot 已从最初的 Funko Pop 玩偶和体育卡牌等小众收藏品扩展至 100 多个类别。目前，女装已跃升为平台第一大品类，其运作逻辑类似于 TJ Maxx 等折扣店，商家借助直播间的高流转能力以深幅折扣快速清仓；高尔夫品类销售额也实现了数倍的同比爆发式增长。在生鲜食品等非传统领域，直播视频直观展示了码头起获的生鲜细节并实时解答疑虑，消除了静态网页上的信任障碍，圣迭戈的海鲜分销商甚至能在单次一小时的试点直播中实现 3000 至 4000 美元的销售额。 平台治理基建与后台赋能型人工智能 随着用户规模达到千万级别，平台的运营安全与信任维护成为核心挑战。LaFontaine 将其信任与安全部门比作超大城市的治安与法律系统，该部门的人员编制占到 Whatnot 全体员工的约 40%。为了在大规模并发交易中控制风险，平台自主研发了基于大语言模型（LLM）的亚秒级规则引擎，实时处理数十亿级数据点。该系统能快速识别直播骚扰、发货延迟或退款率异常，并根据商家的历史信用自动执行警告、暂停或永久封禁。与此同时，直播出镜和高度依赖回头客的商业模式，也对卖家行为形成了天然的信誉约束。 在人工智能应用方面，Whatnot 坚持人际连接是平台的核心价值，因此明确排斥前台虚拟人（AI Avatars）代播模式，认为缺乏真实互动的卖家会迅速流失受众。平台的 AI 战略完全集中于赋能卖家后台运营，利用大语言模型自动化处理繁琐的行政流程，包括智能生成商品列表、推断用于推荐算法的复杂元数据，以及提供经营数据分析，从而帮助真实的人类创业者更高效地运营业务。 全球化布局与高货值合规品类探索 目前 Whatnot 已在包括美国、加拿大、英国等 5 个欧洲国家、日本和澳大利亚在内的 10 个国家开展业务。LaFontaine 指出，尽管各地存在文化差异，但中小企业在收藏品、时尚和数码等品类的商业驱动逻辑高度一致，验证了直播购物作为一种商业媒介跨越地域的持续生命力。 面向未来 12 到 36 个月的规划，Whatnot 的战略重点除了降低运费、改善商品推荐与拓展新国际市场外，还将重点进军高单价与受监管的品类。其中，汽车直播销售被列为关键拓展方向，买家可通过实时全景绕车检查、收听发动机运转声音来评估车况；此外，啤酒、葡萄酒和烈酒等合规酒类商品也将逐步上线，平台力求通过对卖家体验的持续投入来撬动更广泛的消费场景。 虽然 Whatnot 在收藏品和折扣女装等品类展示了强劲的交易额扩张能力与用户粘性，但其能否将直播电商的成功经验顺利复制至汽车、酒类等重度依赖供应链与强监管的高门槛行业，并克服欧美消费者长期形成的静态电商使用习惯，仍有待未来市场的长期检验。

## Background 本期内容由风险投资机构 a16z 于 2026 年 8 月 20 日发布在 YouTube 平台。节目中 a16z 普通合伙人 David George 深度对谈 Whatnot 联合创始人 Grant LaFontaine，探讨了平台如何从早期的宝可梦卡牌与潮玩收藏品起步，逐步演变成用户日均停留 95 分钟、集发现、娱乐与社区于一体的全球领先直播电商生态。 ## TL;DR 节目系统梳理了 Whatnot 突破传统“精准搜索目的型”货架电商、打造“移动端数字购物中心”的商业底层逻辑，揭示了超过 80% 的日活用户即便不产生购买也乐于在直播间社交互动的独特黏性。Grant 分享了从东京酒吧醉酒抢注域名、以 0.5 枚比特币买下核心资产的创业源起，以及平台从潮玩跨越至女装、码头生鲜、高尔夫乃至筹备汽车等 100 多个品类的扩张轨迹。同时，双方深入探讨了赋能鲜活小微实体的供给侧飞轮、投入 40% 人力研发亚秒级风控规则引擎的信任治理，以及坚守真人温度、仅将 AI 用于后台提效而非虚拟人带货的技术边界。 ## Atmosphere & style 整场对谈氛围轻松自如、充满创业者的坦率与幽默，既有聊到童年倒腾全息吉利蛋卡和酒后抢注域名的生动轶事，也有深夜网购码头海鲜、自嘲因关节问题转战高尔夫的生活化调侃。两位对话者在投资人与一线实干家的视角切换中张弛有度，将深刻的零售经济学拆解与严谨的城市级风控治理融为一体，传递出极具感染力的极客热情与商业穿透力。 ## Bulletin - [20:04](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1204s) Whatnot 活跃用户日均停留时长高达 95 分钟，且在任意一天中超过 80% 的用户完全不进行购买，平台实质上已被重塑为一个以内容社交和闲逛探索为核心的“数字购物中心”。 - [23:42](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1422s) 平台头部商家年交易额已突破 1 亿美元且维持 20% 至 40% 以上的健康 EBITDA 利润率；Whatnot 平台累计成交规模已超 80 亿美元，品类从潮玩卡牌迅速横向破圈至女装、生鲜和高尔夫等 100 多个垂直领域。 - [32:29](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1949s) 全公司约 40% 的员工投入在信任与安全（T&S）领域，团队自主研发了亚秒级响应的风控规则引擎并结合 LLM 实时治理音视频违规，以类似运营千万级数字大都会的机制捍卫交易信任。 - [37:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2220s) 在 AI 应用战略上，Whatnot 明确拒绝 AI 数字人替身主播，将技术重心严格限定于后台商品一键上架、推流元数据智能标注与卖家经营分析等提效场景，坚决保留真人连接的温度。 ## Key takeaways - [12:41](https://www.youtube.com/watch?v=XqEr7hk89HY&t=761s) 意图搜索 vs 体验发现：传统电商依赖确定性的搜索框买单，而 Whatnot 还原了线下商场漫无目的闲逛的乐趣，以“模糊意图 + 偶遇好物”的模式打造了极高黏性的娱乐消费体验。 - [09:02](https://www.youtube.com/watch?v=XqEr7hk89HY&t=542s) “用户不在乎宏观大盘”的认知优势：创业早期团队对亚洲万亿美元直播带货大盘或传统 QVC 模式几乎毫无概念，这种“无知”反而帮助团队避开渐进式微调的陷阱，纯粹基于藏家在社交媒体生硬拼凑交易的实际痛点倒逼出产品创新。 - [18:01](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1081s) 需求扩张效应（Demand Expansionary）：直播电商并非存量线上转移，而是类似于当年大型购物中心的崛起，通过内容互动激发消费者此前未曾意识到的潜在购买欲，从根本上放大了零售市场的整体需求规模。 - [21:18](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1278s) 颠覆传统电商的“去中心化”商户逻辑：与亚马逊抹平商家面孔、追求极致算法比价不同，Whatnot 将镜头前有血有肉的中小店主重新推向生意核心，建立起深厚的人格信任与老客高频复购壁垒。 - [15:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=900s) 供给侧零边际成本的降维打击：相比纽约第五大道数百万美元年租金且受制于本地客流的实体门店，Whatnot 商家仅需投入 0 美元开店成本即可面向全球受众全天候直播出货，单位经济模型具备压倒性优势。 - [23:42](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1422s) 收藏品从小众亚文化跨入主流消费资产：随着年轻群体可支配收入提高与怀旧情绪发酵，稀有卡牌潮玩已被视为当代艺术品配置，推动平台大卖家规模天花板达到年销数千万甚至上亿美元。 - [27:07](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1627s) 实体零售的“沙发化”迁移法则：凡是在现实世界中能够支撑起一家独立零售店的品类（如生鲜海产、高尔夫装备），都能在 Whatnot 上运转起来；目前女装更凭借类似 TJ Maxx 的过季折扣走量模式跃升为平台出货量最大的类目。 - [30:30](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1830s) 实时音视频击穿重度决策品类的信任壁垒：圣迭戈海鲜分销商码头直播的爆发证明，主播现场讲解与即时互动能够彻底消除传统图文生鲜网购的安全疑虑，帮助本土小商家直接将生意做向全美。 - [29:27](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1767s) 全球化扩张验证跨文化底层规律：在进驻美、加、欧、日、澳等 10 个国家后，团队发现不同文化背景下的买家黏性与中小商家成长动能惊人一致，印证了直播电商交互形态的普适性。 - [32:29](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1949s) 亚秒级风控规则引擎与严密城市治理：面对平台数千万级用户生态，团队自研能在不足 1 秒内吞吐数十亿数据点的 Rules Engine，结合大模型自动化识别违规行为并执行降权处置，守住生态底线。 - [37:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2220s) 明确 AI 落地边界，严守人际连接：全面拥抱 AI 以解决商品快速上架、推流标签提取与商户数据诊断等后台繁杂任务，但坚决抵制虚拟人带货，确保平台始终建立在人与人真实互动的基础上。 - [39:34](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2374s) 未来路线图：品类深度扩张与物流降本：未来 12 到 36 个月将积极筹备二手车/汽车高清直播拍卖以及酒水类目，同时持续通过规模化履约合作降低买家运费门槛，深化国际化市场渗透。 ## Detailed walkthrough ### [00:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=0s) 传统电商的局限与直播电商的崛起 现代传统电商的核心逻辑建立在强烈的明确意图之上——用户必须清楚知道自己想买什么，通过搜索框输入精准关键词才能完成交易。然而，海外的商业形态正在经历一次重大范式转移。对比来看，直播电商在亚洲（尤其是中国）已经占据了整个电商零售额的 30% 到 40%，几乎是三分之一强；而在美国，这一渗透率目前仅处于微不足道的个位数。 早期欧美的尝试之所以举步维艰，是因为整个链条极其割裂：交易集成度低、物流体系无法打通、支付环节摩擦重重、商品发现缺乏乐趣、售后客户服务（CX）体验糟糕。Whatnot 看到的就是这样一个充满破损环节但潜力巨大的空白市场。不同于传统货架电商，Whatnot 将重心放在了娱乐体验与用户粘性上——用户并不在乎宏观商业概念，他们只关心平台是否好玩、是否有价值。这种以乐趣为驱动的体验创造了惊人的黏性，平台用户如今平均每天停留长达 95 分钟。 ### [01:06](https://www.youtube.com/watch?v=XqEr7hk89HY&t=66s) 全息宝可梦卡片：创始人的商业启蒙 Whatnot 联合创始人 Grant LaFontaine 的电商基因可以追溯到小学三四年级（大约七至九岁）。当时全美掀起了一阵狂热的宝可梦（Pokémon）卡牌浪潮。Grant 像很多同龄孩子一样，拉着父亲去卡牌店拆包，把珍贵的喷火龙（Charizard）等稀有卡片整齐收在卡册里，甚至和小伙伴在街角湖边聚会时尝试“精明地”倒换卡牌。 在 eBay 刚刚兴起、雅虎拍卖（Yahoo Auctions）与其分庭抗礼的年代，Grant 偶然发现了网上交易卡牌的可能：“等一下，我玩宝可梦卡还能赚钱？我这么喜欢它，居然能靠它赚钱！”在父亲的协助下，他注册了账号，并上架了自己的闪卡（holographic cards）。他的第一笔在线交易是一张售价 10 美元的吉利蛋（Chansey）全息卡片。在那个缺乏线上支付基础设施和网络信任机制的互联网洪荒时代，远方的买家甚至需要去邮局购买汇票寄到他家，年幼的 Grant 收到纸质汇票后去邮局兑现，再仔细打包卡片寄回给买家。这段早期经历成为了 Whatnot 的真正源头。回溯 Grant 的职业履历：童年倒腾宝可梦卡，随后在 YouTube 长期深耕视频领域，后来又加入 Facebook 负责 Marketplace（二手市场交易业务）——Whatnot 正是他一生核心技能与爱好的终极集大成者。 ### [03:49](https://www.youtube.com/watch?v=XqEr7hk89HY&t=229s) 东京居酒屋的醉话与 0.5 枚比特币买下的域名 关于 Whatnot 的诞生，Grant 和另一位联合创始人 Logan Head 各自有着不同的记忆版本，两者却拼凑出了完整的创业图景。Logan 记忆中的源头发生在 2019 年 1 月的新年假期。两人当时在东京的一家酒吧里喝得酩酊大醉，由于各自此前都有过创业经历，几杯酒下肚便借着酒劲热烈讨论未来想做的大事。他们当晚就着醉意畅想搭建一个新型交易市场平台，一口气在 GoDaddy 上抢注了三四十个域名。然而旅行一结束，回到现实生活的繁琐工作中，这个点子便搁置了整整七八个月。 到了 2019 年 10 月，随着 Logan 即将结束手头的工作，Grant 翻出了当时在东京喝醉时买下的域名清单，提议是时候正式行动了。他们敏锐地观察到身边的朋友们开始回流收藏品圈子——潮玩泡泡玛特/方头公仔（Funko Pops）、漫画书、球星卡等品类热度暴涨。伴随着智能手机长大的这一代年轻人逐渐拥有了可支配收入，老旧的收藏品交易模式却已停滞多年。他们断定：收藏品加上社交电商（Social Commerce）必定大有可为。 ```mermaid flowchart TD accTitle: Whatnot 创业与命名历程 accDescr: 描述从 2019 年东京酒吧醉酒讨论到正式上线 Funko Pop 交易平台的关键节点。 subgraph Ideation ["灵感与冲动期 (2019 初 - 2019 秋)"] A["2019.01 东京酒吧喝醉<br/>头脑风暴并抢注 30-40 个域名"] --> B["搁置数月，各自在职"] B --> C["2019.10 洞察收藏品回潮趋势<br/>发现移动端社交电商机会"] end subgraph Execution ["执行与命名风波 (2019 年底)"] C --> D["2019.12 双双离职<br/>全职创办公司"] D --> E["Grant 定下临时名: Always Legit<br/>(只求快速推进，不纠结品牌)"] E --> F["设计师 John 强烈抗议<br/>1 小时头脑风暴筛选 10 个候选名"] end subgraph Acquisition ["域名收购谈判"] F --> G["锁定优质域名: whatnot.com"] G --> H["原持有人要价 1 枚比特币 ($10,000)"] H --> I["Logan 极限砍价至 0.5 枚比特币 ($5,000)<br/>中间甚至乌龙错发 whynot.com"] I --> J["正式以 Whatnot 启动<br/>上线正品认证 Funko Pop 交易平台"] end ``` 2019 年 12 月，两人双双辞职，全职投身创业，首发业务定位为经过鉴定的 Funko Pop 二手交易市场。当时 Grant 本着“野蛮推进、产品为王”的心态，将公司临时命名为极度草率的“Always Legit”。这一名字遭到了平台第一位设计师 John 的坚决抵制，直言“绝不能和这么难听的名字产生关联”。 于是三人被逼进会议室展开了一小时的高效头脑风暴，筛选出 10 个彼此都认可的候选名分头联系收购。其他名字要价动辄 10 万美元以上，而没有获得任何融资的他们根本负担不起。幸运的是，Logan 联系上了 `whatnot.com` 的持有人。对方开价 1 个比特币（当时一枚比特币价值约 10,000 美元），早期就持有比特币的 Grant 与 Logan 联手还价，最终硬是将价格砍到了 0.5 个比特币（折合仅 5,000 美元）。其间还发生了一个搞笑的乌龙插曲：由于转接托管平台对加密货币操作不熟，对方甚至一开始错把 `whynot.com` 当成了目标域名发了过来。但最终，他们以极低成本锁定了这个寓意“包罗万象、各类物件”的完美名字，为日后扩展到全品类电商打下了基石。 - **关键节点与数据**： - Grant 第一张在线卖出的卡牌：全息吉利蛋卡（Holographic Chansey），成交价 10 美元。 - 目前 Whatnot 活跃用户的日均使用时长：约 95 分钟。 - 核心创始人辞职创业节点：2019 年 12 月。 - 域名收购成本：以 0.5 枚比特币（时值 5,000 美元）买下 `whatnot.com`。 ### [09:02](https://www.youtube.com/watch?v=XqEr7hk89HY&t=542s) “用户根本不在乎什么大盘市场”：误打误撞的认知红利 两人的对话从 Whatnot 域名的最终敲定切入。Grant 透露，当时他们花了足足 36 个小时才终于联系上域名的持有者并成功买下；在此之前，他们甚至一度认真考虑过使用“Why Not”这个备选名称。David 笑着感叹这个小插曲十分有趣，并顺势将话题引向 2019 年底 Whatnot 刚起步时的外部市场环境：当时在美国，Facebook Marketplace 凭借显著的网络效应风头正劲；在亚洲，直播电商（Live Commerce）作为一种全新的独立商业形态已初露锋芒；而在西方主流零售语境中，大家对“屏幕购物”的全部认知基本还停留在 QVC 这类传统电视购物频道。 面对 David 关于当时创业底气从何而来的提问，Grant 坦诚地表示，在 2019 年 12 月（即新冠疫情暴发前夕），他们对宏观市场格局其实“根本谈不上有什么高深的认知”。他们当时压根不知道亚洲已经跑通了万亿美元级的直播带货模式；至于 QVC，Grant 的第一反应则是：“谁会想在 QVC 上买东西啊？那都是老奶奶（grannies）才去的地方，跟我同龄的人绝不可能去 QVC 购物。”当时收藏品领域虽然存在一些老牌交易平台，但整个赛道给人的感觉非常陈旧死板。David 对此评论道，对所谓大盘格局的“无知”反而是极大的资产而非劣势——如果创业团队一开始就熟知既有市场，往往会不由自主地陷入对现有模式做局部渐进式改良的陷阱，而无法创造全新的物种。 Grant 对此表示“一千个百分点地赞同”。他一针见血地指出，许多创业公司热衷于对标亚洲万亿美元的庞大市场，反而容易被宏观概念带偏，因为“用户根本不在乎什么大盘市场（users don't give a **** about the market）”。消费者不会因为某个行业被称为“直播电商”就产生买单冲动，他们只在乎产品能否带来快乐与切实的价值。当时 Grant 和团队正专注于搭建 Funko Pop 潮玩交易平台，每天直接与深度买家泡在一起。他们敏锐地观察到：许多核心用户正在用极其原始的方式“魔改”社交媒体平台来进行直播卖货，但那种体验完全是破碎的——交易链路没有打通、物流无法追踪、支付极其繁琐、缺乏商品发现机制、售后与客户体验（CX）更是一团糟。意识到这一真实痛点后，联合创始人 Logan 将自己关在房间里闭关开发了数周，拿出了 Whatnot 的首个直播版本。Grant 随即亲自开播，所有上架商品瞬间被一抢而空，整个平台的发展篇章由此正式掀开。 - **域名波折**：在耗时 36 小时拿下 `whatnot.com` 之前，团队险些采用“Why Not”作为品牌名。 - **早期盲区与误解**：2019 年底起步时团队完全不了解亚洲直播带货的繁荣，对欧美电视购物的印象仅停留在“老奶奶才会用的 QVC”。 - **痛点驱动创新**：并非照搬海外成熟模式，而是发现 Funko Pop 潮玩藏家在社交媒体上生硬拼凑直播交易，支付、物流与商品发现体验全面脱节。 - **快速原型验证**：联合创始人 Logan 闭关数周手搓出 V1 核心产品，Grant 首次直播便实现全部售罄。 --- ### [12:41](https://www.youtube.com/watch?v=XqEr7hk89HY&t=761s) 意图搜索 vs. 体验发现：把“数字购物中心”搬进手机 谈及 Whatnot 为用户创造的核心价值时，David 抛出了一个电商领域的本质问题：传统线上购物通常始于极其强烈的购买意图（Search Intent），那么 Whatnot 的产品底层究竟更偏向“纯娱乐”还是“意图消费”？平台买家的行为分布又是怎样的？ Grant 给出了一记精妙的类比：逛 Whatnot 的心智模型，本质上与线下逛购物中心（Shopping Mall）如出一辙。在传统电商（如早期 Amazon 或 eBay）中，消费路径是高度意图导向且功能化的：用户明确知道自己要买什么特定型号，输入搜索框、下单结算，流程缺乏探索（Discovery）的乐趣，更谈不上体验与娱乐。而线下逛街则是另一套逻辑——消费者心里可能只有一个模糊的方向（例如 Grant 提到自己最近想买双新运动鞋），但并不知道确切的款式与品牌；走进商场漫无目的地闲逛半小时，在浏览和互动中偶然遇上心仪的好物，最后满载而归。Whatnot 还原的正是这种“模糊意图 + 发现式惊喜”的体验。 ```mermaid flowchart TD accTitle: 传统电商与 Whatnot 体验模式对比 accDescr: 对比传统意图驱动电商与 Whatnot 基于社区闲逛探索的双轨驱动路径 subgraph Traditional ["传统意图型电商 (Amazon / eBay)"] direction TB A["明确购买意图<br/>(精准搜索目标)"] --> B["比价与功能筛选"] B --> C["即时结账离开<br/>(无停留与娱乐)"] end subgraph Whatnot ["Whatnot 数字购物中心模式"] direction TB D["模糊需求或无目的登录"] --> E["进入直播间: 观看、聊天、围观"] E --> F{"是否产生消费?"} F -- "绝大多数日常" --> G["社区社交、围观竞拍、纯享娱乐"] F -- "偶遇心动好物" --> H["互动出价、拼单捡漏、完成购买"] end ``` 针对 David 列举的平台功能——观看、打字聊天、出价竞拍、购买、退货，Grant 强调，这种多维度的互动结构正是 Whatnot 具有极高粘性的根源所在。用户在平台上的动机千差万别：有人是为了寻找超值折扣，但更多人纯粹是为了获取内容消费体验与社交娱乐。Grant 披露了一个惊人的平台数据：**在任意给定的一天中，Whatnot 上的绝大多数用户根本不会发生任何购买行为**。他们登录平台只是为了挂在直播间看主播开箱、在公屏打字聊天、围观竞拍过程。这种氛围就像童年时代与死党相约去商场闲逛打发时间一样，大部分时候并不为了买具体的东西，只是享受沉浸其中的社交与现场感。 --- ### [15:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=900s) 供给侧的颠覆性价值与长尾电商的万亿空间 David 进一步将话题延展至市场宏观天花板。数据显示，亚洲直播电商占整体零售电商的比例已高达 30% 到 40%，而当下的美国市场渗透率仍处于单薄的个位数（Single Digits）。Grant 坚定地认为，这一巨大的渗透率鸿沟在未来必然会迅速弥合，核心原因在于直播电商为商家提供了远比传统商业更丰富、更高维且更具效率的经营杠杆。 为了形象地说明供给侧的压倒性优势，Grant 做了一个鲜明对比：如果一个时尚服饰品牌想要在线下开辟业务，选址在纽约第五大道这样的黄金商业街，光是每年支付的租金就高达数百万美元，且每天只能覆盖朝九晚五路过纽约实体店的极少数本地客流；即便实体店能提供触摸质感与面对面交流的绝佳体验，其物理瓶颈也显而易见。而在 Whatnot 上，任何优质商家只需投入“0 美元”的开店边际成本，就能立刻面对一个不受地理限制、完全向全球敞开的无边界消费受众。这种对供给侧商家而言极度诱人的单位经济模型，构成了平台飞轮的根本动能——零门槛、高收益的供给生态自然会吸引海量多元商户涌入，而丰富生动的供给侧内容则会源源不断地虹吸海量消费需求。 ```mermaid flowchart LR accTitle: 实体零售与 Whatnot 供给侧模型对比 accDescr: 对比第五大道线下精品店与 Whatnot 直播间的成本及受众覆盖规模 subgraph Physical ["传统线下模式 (如纽约第五大道)"] P1["高昂固定成本<br/>(数百万美元年租金)"] --> P2["物理受众受限<br/>(本地客流 / 朝九晚五)"] end subgraph WhatnotSupply ["Whatnot 供给侧优势"] W1["零边际开店成本<br/>(0 美元入驻开播)"] --> W2["无边界全球受众<br/>(全天候流量聚合)"] end Physical -.->|效率升级与替代| WhatnotSupply ``` 针对外界常年对 Whatnot 抱有的质疑——“作为长尾非标品平台，它的天花板顶多也就是下一个 Etsy 或 eBay”，David 给出了强有力的反驳。他指出，目前美国整体电商大盘规模已达约 1 万亿美元（与亚洲直播电商体量相当）。十年前，eBay 曾占据美国电商大盘约 15% 的份额；虽然如今 eBay 在整体电商中的份额已稀释至不足 1%，但其年度交易额（GMV）依然稳固在 800 亿美元的量级。这充分证明，并不是所谓的非标长尾需求变小了，而是整个电商盘子在以超乎想象的速度膨胀。哪怕仅以当下 eBay 的存量体量作为参照锚点，长尾与非标品交易本身也至少是一个 800 亿美元的基本盘；而对于正在用直播与娱乐重构消费体验的 Whatnot 而言，用过去陈旧的电商天花板来设限显然远远低估了它的长期潜力。 - **渗透率断层**：中国直播电商占电商总规模约 30%~40%，而美国仍处于低个位数阶段，存在巨大的结构性追赶空间。 - **第五大道 vs. 零成本开播**：数百万美元租金 + 局部客流限制的线下实体模式，正在被零成本门槛 + 全球无边界流量的直播供给侧彻底降维打击。 - **万亿美元大盘与破除 eBay 迷思**：美国整体电商规模约为 1 万亿美元；eBay 份额即便从 15% 跌破 1%，体量仍有 800 亿美元之巨，有力印证了非标长尾电商绝非小众生意，且伴随大盘扩张依然具备极高的天花板。 ### [18:01](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1081s) 摆脱“eBay 标签”与直播电商的需求扩张效应 主持人 David George 提出，电商市场规模庞大（在 800 亿美元到 1.2 万亿美元之间），由于直播购物体验远优于传统模式，Whatnot 能够从中蚕食并创造多大的市场份额。对此，Grant LaFontaine 明确表示，外界往往因为 Whatnot 最初做收藏品（Collectibles）起家而将其与 eBay 锚定，但事实上**直播形态在底层逻辑上截然不同**。eBay 早期以个体小黄牛或二手倒手转售者（Resellers）为主，而 Whatnot 平台的主力则是实体小微企业（Small Businesses）乃至正规品牌方。这些商家平时可能拥有实体店或运营独立的 Shopify 独立站，传统意义上绝不会入驻 eBay。 更核心的差异在于需求端的扩张。传统电商是高度以“搜索和目的”为导向的（Intent-driven），消费者必须明确知道自己要找什么，本质上只是把线下的既有交易以更高效率搬到了线上；而直播电商具有极强的**需求扩张性（Demand Expansionary）**，用户在围观与互动中会发现自己原本根本不知道或没想过会感兴趣的新商品。David George 用百货公司和现代购物中心的诞生打比方：它们当年并不是简单分流了街边小摊，而是从根本上重塑并放大了整个市场的消费需求。直播电商未来不仅有机会占据整体零售市场 30% 到 40% 的份额，还将促使整个市场的总量天花板大幅提升。 ```mermaid flowchart TD accTitle: 传统电商与 Whatnot 直播电商的交易模式对比 accDescr: 对比传统目的型电商与 Whatnot 娱乐发现型直播电商在商家、逻辑和需求效应上的核心差异 subgraph Traditional["传统目的型电商（如 eBay / Amazon）"] T1["商家定位: 匿名转售商 / 标品铺货"] --> T2["购买逻辑: 确定性搜索（已知需求）"] T2 --> T3["市场效应: 存量迁移（单纯线下转线上）"] end subgraph Whatnot["Whatnot 直播电商生态"] W1["商家定位: 品牌店 / 实体小商家 / 达人"] --> W2["购买逻辑: 直播互动与内容娱乐"] W2 --> W3["市场效应: 发现未知乐趣与需求扩张"] end ``` ### [20:04](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1204s) 每日 95 分钟与超过 80% 的“不买纯逛” David 将话题转向 Whatnot 令人震惊的用户黏性数据：用户在平台上的日均使用时长高达 **95 分钟**。这一数字不仅打破了电商平台的常规天花板，甚至直接比肩头部社交与娱乐媒体应用。 Grant 解释道，如果将平台对买家的核心价值主张简化为一个词，那就是“好玩（Fun）”。长期以来，在线网购一直极度缺乏乐趣，而曾经承载社交逛街功能的线下购物中心也逐渐失去了吸引力。在 Whatnot 上，用户不仅仅是为了下单而来：在任意一天中，**超过 80% 的在线用户完全没有进行任何交易**，他们仅仅是坐在屏幕前观看直播、在公屏打字聊天、参与社区互动并感受纯粹的快乐。David 对此深有共鸣，表示自己也经常在上面当作纯娱乐节目看很长时间，买东西反而成了顺带的副产物。这种“娱乐消遣与新奇探索并存”的特质，构成了平台最坚固的用户护城河。 ### [21:18](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1278s) 赋能实体小微企业：反向颠覆传统电商的“去中心化” 在谈到供给侧时，双方深入探讨了 Whatnot 如何重塑中小商家的生存空间。过去二十到三十年间，以亚马逊（Amazon）为代表的中心化电商平台为了追求极致履约效率，普遍采取“抹除商家面孔”的做法：平台通过复杂的算法竞价、自营与第三方混战、跨境标品拼刺刀等高度功利化的机制，将具体的人和品牌隐匿于商品详情页与运送箱之后。 Whatnot 的哲学则恰恰相反，它让镜头前鲜活的店主重新成为生意的中心，如同“街角小酒馆的老板”或熟人社区店店主： - 商家在平台上建立属于自己的独立直播间、品牌形象与忠实客群，顾客会因为喜欢主播的个性和选品而反复光顾。 - 平台培育的经营规模极广，从夫妻档两人团队，到员工规模达数百人的成熟商业组织。 - Whatnot 为中小企业创造了数以十亿美元计的成交额，让有创意、有野心的创业者无需陷入无休止的比价内卷，也能脱颖而出。 ### [23:42](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1422s) 收藏品出圈主流文化与年销破亿美元的超级卖家 对于最初起家的收藏品（卡牌、潮玩等）品类，Grant 指出这一趋势在过去六年半中持续爆发且未见衰减。早年 Whatnot 抓住的是同龄人因怀旧情感和可支配收入增加而重返童年爱好的红利；而如今，收藏品已经彻底从所谓的“小众亚文化（Niche）”蜕变为席卷主流流行文化的核心资产。不仅各界顶级明星、退役与现役运动员下场参与，高净值人群也开始将罕见卡牌视为当代艺术品进行资产配置——过去富裕阶层热衷收藏大师名画，现在的年轻人则斥资数百万美元争夺像勒布朗·詹姆斯（LeBron James）1 of 1 折射卡（Refractor）这样的顶尖藏品。 随着市场主流化，平台商家的天花板被大幅推高： - 平台的头部大卖家年 GMV 已经能够突破 **1 亿美元**，并保持着极为健康的利润表现，EBITDA 利润率高达 **20%、30% 乃至 40% 以上**。 - Whatnot 自身在去年及今年至今的交易额均已突破 **80 亿美元**，但平台坚持只抽取一小部分服务费，绝不随意提价或压榨利润。 - 公司的底层信条是“把绝大部分价值留给商家”，让平台伴随卖家的成长而自然增长。 ### [26:18](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1578s) 品类全线破圈：突破 100+ 品类与“码头生鲜” 在巩固卡牌收藏品统治地位的同时，Whatnot 现已成功横向扩张至 **100 多个细分品类**。除了已经成为主力增长引擎的时尚服装（Fashion）赛道外，平台正在不断进军各类新奇的实体商品领域。 Grant 特别分享了他们近期大力扶持的生鲜食品与饮料（Fresh Food & Drink）品类。例如，有渔业商家直接把直播间搬到圣地亚哥（San Diego）的码头现场，刚靠岸卸货就当场展示售卖野生海鲜，生猛的现场感引发了用户的狂热抢购。David 随即笑着打岔透露，自己和 a16z 联合创始人 Ben Horowitz 某次吃完晚饭后，也忍不住跑到 Whatnot 上看直播下单买生鲜海产。此外，类似地方风味烤肉（Barbecue）等具有强烈烟火气和展示属性的美食类目，也正在直播间里催生出前所未有的消费体验。 ### [27:07](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1627s) 生鲜食品的起步与线下商场的“沙发化”迁移 谈到平台的品类扩张，Grant 笑着分享了自己的日常体验：他之前在平台上的一家餐厅买了一大批美味的烧烤，现在还塞满了家里的冰箱。生鲜食品（Fresh Food）虽然在 Whatnot 上才刚刚起步，但他对此异常兴奋，因为美国传统的食品供应链极其不透明，而直播能让消费者清楚了解食物产自何处，听一听食物背后的真实故事。紧接着，Grant 话锋一转聊起了高尔夫——他调侃道，不知为何身边的朋友全都开始打高尔夫了，David 打趣说“那是因为你开始变老了”，Grant 也顺着自嘲由于自己髋关节不好，其他剧烈运动做不了，正好赶上这波潮流。如今 Whatnot 上的高尔夫品类迎来爆发式激增，年同比增长达到数百个百分点，大量专业高尔夫用品店纷纷入驻直播。 David 总结道，任何属于“重度决策”（highly considered）的商品天然都非常契合 Whatnot，因为这类产品极度需要详细的解说、鲜明的人格魅力以及背后的故事支撑。Grant 对此进一步提炼出一个通用的零售法则：凡是现实中有人愿意为其开设线下实体零售店的商品，就一定能在 Whatnot 上运转起来。不管是专门店还是传统大型购物商场（Shopping Mall）里的柜台，既然用户以前愿意专程驱车去逛，现在他们完全可以直接蜷在客厅沙发上，捧着手机完成同样的探索与选购。 - 生鲜食品痛点：传统零售供应链极不透明，缺乏产地溯源与生产故事。 - 高尔夫品类：受众年龄层契合，年增速达到数倍（数百个百分点），大量实体球具店转战线上直播。 - 选品通用逻辑：“凡是能支撑一家独立零售店的品类，都能在 Whatnot 上做成”，核心在于把线下商场的逛街体验搬上移动端。 --- ### [28:24](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1704s) 进军服饰领域：用直播打通折扣清仓与走量飞轮 当话题转到服装与服饰穿搭时，Grant 透露了一个令人惊讶的数据：女装目前已经一跃成为 Whatnot 平台交易量最大的单项品类。平台上的商品形态非常丰富，但占据主导地位的模式实际上极度贴近知名折价零售商 TJ Maxx 以及大牌折扣鞋店 DSW——主打品牌过季库存（Old Season Inventory）和极具杀伤力的价格折扣，消费者经常能在上面淘到大牌尖货。Grant 回忆起自己童年时去街头平价鞋店花 40 美元淘一双耐克鞋的兴奋感，指出在购物这门生意里，没有什么比“捡到了大便宜”更能打动买家。 这种折价模式与直播形态产生了极强的飞轮效应。服装库存的痛点在于去化周期，而直播间高频、快节奏的互动能够短时间内带动极大的出货量（Volume）。正是因为能够帮助卖家实现规模化出货，商家才愿意在直播中割肉给出极低的折扣，这反过来又为买家创造了传统电商货架难以比拟的极致性价比。 --- ### [29:27](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1767s) 全球化扩张：10 国布局与高度一致的底层动态 在攻克美国市场后，Whatnot 迅速将版图拓展至海外。目前平台已进驻约 10 个国家：北美涵盖美国和加拿大；欧洲开通了 5 个核心国家；亚太地区则成功落地日本与澳大利亚，初步完成了全球主要消费市场的拼图。 Grant 坦言，国际化扩张中最令人意想不到且反直觉的发现，是不同国家的底层运转逻辑竟然惊人地一致。无论是在欧美还是亚太，团队此前在美国帮扶起的一批小微商家，在海外同样能够以相似的轨迹起步并建立生意。从收藏品、时尚服饰到消费电子产品，各个市场的买家行为、卖家成长动力与互动黏性几乎如出一辙。这一现象让团队更加坚信，直播带货绝非某种特定文化环境下的短期产物，而是一种具备极强生命力和持久价值的消费交互形态。 --- ### [30:30](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1830s) 圣迭戈海鲜批发商的故事：视频如何击穿生鲜信任壁垒 David 提到了一个 Grant 曾分享过的生动案例：一位圣迭戈的海鲜分销商在 LinkedIn 上主动私信了 Grant。在接洽后的第二周，这位批发商就开启了首场试点直播，仅在测试阶段的短短几小时内就斩获了三四千美元的鲜鱼销售额。 对于这个令人惊叹的冷启动案例，Grant 解释称，直播的杀手锏在于“发现感”与“建立极致信任”。这位分销商深谙鱼类生意的门道，常年精选圣迭戈当季最地道、最新鲜的海产。直播间里，他会拿起观众闻所未闻的小众海鱼，娓娓道来它背后的可持续捕捞标准，告诉大家这条鱼今天清晨刚从码头卸货，甚至如数家珍地介绍出海捕鱼的渔民。这种身临其境的体验，让 Whatnot 庞大的买家池迅速转化为实际购买力，轻松做到每小时数千美元的成交额。 David 指出，如果没有直播，这类本土小商家想要做全美生意几乎是不可能的任务，因为他们最多只能辐射本地圈子。而网购生鲜最大的心理阻碍是“食品安全”——普通用户在静态网页上看到生鲜冷冻鱼时，第一反应往往是“这鱼真的新鲜吗？吃了会不会中毒拉肚子？”。而在直播间中，屏幕另一端是一个有血有肉的鲜活个体，买家可以实时向他提问，看他如何在镜头前打包冰鲜箱。视频展现的真实人格彻底打破了传统图文网站“只能听天由命赌运气”的无力感。 --- ### [32:29](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1949s) 规模化运营与风控：亚秒级规则引擎与万人数字大都会 面对平台高速扩张所带来的天量并发和交易洪峰，David 赞赏 Grant 和联合创始人 Logan 的核心特质在于“执行层面的咬住不放”（relentlessness）——在早期还没有强大网络效应壁垒时，团队就对任何可能发生故障的漏洞保持极高的敏锐度。谈到平台规模化后的最大挑战，Grant 将核心归结为两个字：“信任”。 为了守护信任，Whatnot 奉行“度量一切”（measure everything）的硬核原则，并进行了极其沉重的人力与技术倾斜。目前，仅**信任与安全（Trust and Safety）团队就占到了全公司员工总数的约 40%**。在底层架构上，团队自主研发了一套名为“规则引擎”（Rules Engine）的风控中枢系统，该系统能在不足 1 秒（Sub-1-second）的超低延迟内实时吞吐数十亿级的数据点。一旦主播在直播流中出现言语骚扰，大语言模型（LLM）会毫秒级捕获，并联动规则引擎调取该卖家的历史履约画像，在极短时间内自动化执行警告、禁言、下播甚至封号处理；同样，针对发货拖延、退款率异常等供应链指标，系统也会自动实施降权或惩处。 ```mermaid flowchart TD accTitle: Whatnot 信任与安全风控架构流程 accDescr: 展示直播间行为与履约数据如何经由 LLM 和规则引擎进行亚秒级自动化处置 subgraph DataIngestion ["多源信号采集 (数十亿数据点)"] A["直播流音视频与聊天内容"] B["订单履约 (发货时效/退款率)"] end subgraph DecisionEngine ["智能检测与评估"] C["LLM 违规检测与内容理解"] D["卖家历史履约画像与评分权重"] E["Rules Engine 规则中枢\n(Sub-1-second 亚秒级处理)"] end subgraph Enforcement ["全自动分级处置"] F["实时违规: 警告 / 强制下播 / 封禁"] G["履约异常: 业务受限 / 扣除权益"] end A --> C C --> E B --> E D --> E E --> F E --> G ``` Grant 将 Whatnot 现在的生态打了个形象的比方：每天有数千万人涌入平台，这相当于直接在云端管理一两座纽约市或整个东京大都会。在这样庞大体量的城市中，Trust & Safety 团队扮演的角色就是整座城市的警察部队与司法仲裁系统，必须制定法律，并由系统与专员负责执行。尽管庞大的人口基数不可避免会滋生个别灰色事件，但得益于自动化的监控工具以及“全流程直播无死角”的透明特性，任何不良商贩在镜头前都极难蒙混过关。加之绝大多数卖家都立足于长期做生意、依赖老客复购，这种公开透明的直播环境反而从根本上倒逼了卖家提供更优质的服务。 - 团队资源配比：信任与安全（T&S）相关人员占全公司员工的 40%。 - 规则引擎（Rules Engine）：处理延迟在 1 秒以内，实时分析数十亿级数据点。 - 自动化治理范围：从 LLM 识别骚扰言论、直播违规，到履约延迟和退款率偏高，全部实现规则化程序触发。 - 城市治理隐喻：规模相当于两座纽约市的人口体量，依托直播透明化与平台司法体系保持生态底线。 ### [36:35](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2195s) 信任与安全：平台治理如同运营一座超级城市 在探讨平台规模化运营的挑战时，团队将整个公司大约 40% 的人力与资源投入到了信任与安全（Trust and Safety）以及类似“执法”的工作中。运营这种规模的直播电商平台，难度无异于同时掌管纽约市的警察局与立法系统：不仅要在规则制定上与时俱进，还要日夜紧盯层出不穷的新型风险漏洞，持续在关键安全节点上进行重金投入和系统优化。这种高强度的平台治理构成了 Whatnot 业务最坚固的护城河，也直接引出了利用前沿人工智能（AI）来协助风控与执法的探讨。 ### [37:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2220s) AI 的边界：赋能卖家效率，绝不取代真人温度 针对外界关于 AI 与直播电商结合的设想，David George 提出疑问：Whatnot 本质上是一个建立在买家与卖家“人机互动”之外、极度依赖“人与人真实连接”的平台，那么在前端产品与卖家工具中，该如何有效引入 AI，同时又不破坏这种人情味？Grant LaFontaine 明确表示，平台的生态机制本身具有强大的自我修正能力。如果某个卖家异想天开地去开发一个 AI 数字人虚拟主播（AI Avatar）来带货，买家根本不会买账，更不会复购；这种模式在 Whatnot 上完全无法成为一门成立的生意。 Whatnot 的核心精髓在于重现逛线下小店的温暖体验：走入一家熟悉的店铺，认识店主本人，结识经常光顾该店的其他同好买家。正是这种充满欢声笑语和归属感的社交体验吸引着用户不断回访。因此，优秀的卖家绝不会试图抽离真人互动，而是思考如何借助 AI 来提高后台的运营效率： - **商品快速上架（Listing Items）**：借助 AI 视觉与文本能力大幅降低卖家拍摄、录入和标定商品信息的门槛。 - **推流元数据自动推断（Metadata Inference）**：在卖家开启直播时，利用大语言模型（LLM）自动提取并补全复杂的品类与推荐元数据，提升平台的算法分发与发现效率。 - **商业数据分析与经营指导（Seller Analytics）**：通过 AI 深度剖析卖家的销售表现，指出哪些环节做得好、哪些品类滞销、以及可以在哪些关键节点优化毛利。 - **拒绝数字人替身**：明确表态当前绝不会投入精力为卖家提供 AI 替身工具，因为“人们来到 Whatnot，归根结底是为了人与人之间的真实连接”。 ```mermaid flowchart TD accTitle: Whatnot AI 应用哲学与落地边界 accDescr: 展示 Whatnot 如何在后台效率中使用 AI，同时在前台互动中坚决保留真人温度 subgraph FrontEnd ["前台用户体验（坚守真人温度）"] RealHuman["真人主播互动"] Community["同好买家社交圈"] RejectAvatar["坚决不采用 AI 虚拟主播"] end subgraph BackEnd ["后台经营赋能（全面拥抱 AI）"] AI_Listing["智能一键上架与图文解析"] AI_Meta["LLM 自动推断直播推荐元数据"] AI_Analytics["卖家经营诊断与销售分析"] AI_Safety["风控安全与反欺诈审核"] end BackEnd -->|"降低琐碎杂务"| RealHuman RealHuman -->|"营造真实小店氛围"| Community ``` ### [39:34](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2374s) 展望未来：汽车直播拍卖、品类破圈与全球扩张 在展望未来 12 到 36 个月的规划时，Whatnot 的长期路线图依然回归最底层的商业逻辑：帮助越来越多的中小商家在平台上扎根并扩大生意规模。为了达成这一目标，公司将重点推进多品类扩张与更深维度的基础设施建设： Grant LaFontaine 兴奋地透露，自己最渴望上线的一个品类就是**二手车/汽车（Cars）**。在他看来，在 Whatnot 上买车是一件无比自然且令人兴奋的事——买家可以通过高清视频与主播的第一视角，全方位审视车况细节、聆听引擎轰鸣声、感受真实的试驾体验。谈及此，他开玩笑说这可能对自己的银行账户是个巨大的考验，但他显然已经准备好以“吃自己的狗粮”（Dogfooding）为由亲自上手体验一番了。除了汽车，酒类（啤酒、葡萄酒、烈酒）也是平台未来积极筹划的重要垂类。 与此同时，健康的卖家生态必须以极致的买家体验为基础，平台正在从以下几个维度持续发力： - **买家支持与履约降本**：持续重金投入客户支持与信任安全体系，并与物流伙伴深度合作打下运费价格，减轻买家运费负担。 - **供给质量跃升**：不断扩充平台商品品类的丰富度，严格把控正品率与成色质量。 - **跨区域全球扩张**：在立足现有欧美市场的基础上，进一步筹划进军更多国际市场。 正如 David George 总结所言，Whatnot 展现出了对“卖家体验”近乎偏执的狂热关注，正是这种全心全意帮助卖家成功的底层动力拉动了庞大的买家群体；而不断将买家体验打造得更实惠、更有趣、更安心的飞轮效应，正是 Whatnot 彻底颠覆传统电商版图的核心动力。 ## Notable quotes > [01:06](https://www.youtube.com/watch?v=XqEr7hk89HY&t=66s) 等一下，我玩宝可梦卡还能赚钱？我这么喜欢它，居然能靠它赚钱！ — Grant LaFontaine > [09:02](https://www.youtube.com/watch?v=XqEr7hk89HY&t=542s) 谁会想在 QVC 上买东西啊？那都是老奶奶才去的地方，跟我同龄的人绝不可能去 QVC 购物。 — Grant LaFontaine > [09:02](https://www.youtube.com/watch?v=XqEr7hk89HY&t=542s) 用户根本不在乎什么大盘市场（users don't give a **** about the market） — Grant LaFontaine > [12:41](https://www.youtube.com/watch?v=XqEr7hk89HY&t=761s) 在任意给定的一天中，Whatnot 上的绝大多数用户根本不会发生任何购买行为 — Grant LaFontaine > [20:04](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1204s) 好玩（Fun） — Grant LaFontaine > [23:42](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1422s) 把绝大部分价值留给商家 — Grant LaFontaine > [27:07](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1627s) 凡是现实中有人愿意为其开设线下实体零售店的商品，就一定能在 Whatnot 上运转起来。 — Grant LaFontaine > [32:29](https://www.youtube.com/watch?v=XqEr7hk89HY&t=1949s) 度量一切（measure everything） — Grant LaFontaine > [37:00](https://www.youtube.com/watch?v=XqEr7hk89HY&t=2220s) 人们来到 Whatnot，归根结底是为了人与人之间的真实连接 — Grant LaFontaine ## Mentioned Whatnot, Grant LaFontaine, Logan Head, David George, a16z, Ben Horowitz, eBay, Amazon, Shopify, Facebook Marketplace, QVC, Etsy, YouTube, GoDaddy, TJ Maxx, DSW, Funko Pops, Pokémon, 喷火龙 (Charizard), 吉利蛋 (Chansey), 勒布朗·詹姆斯 (LeBron James) ## Video description a16z General Partner David George is joined by Grant LaFontaine, co-founder of Whatnot, to unpack how a marketplace that started with collectibles evolved into one of the world's leading live shopping platforms. Grant traces the company's origins from selling Pokémon cards online as a kid to discovering live commerce by watching Whatnot's earliest customers hack together sales on social media. They discuss why Whatnot thinks less like a traditional e-commerce marketplace and more like a digital shopping mall, where discovery, entertainment, community, and commerce all happen at once. Today, users spend roughly 95 minutes a day on the platform, and most aren't even buying something on a given day. They also explore how Whatnot is enabling small businesses to reach global audiences, expanding from collectibles into categories like fashion, food, and golf, and using AI to make sellers more efficient without replacing the human connection at the center of the experience. Timestamps: 00:00 - Intro 01:06 - The Holographic Pokemon Card That Started It All 03:49 - How Grant & Logan Settled on Whatnot (Drunk in a Tokyo Bar) 09:18 - The State of the Market: Facebook Marketplace, China & QVC 12:41 - Entertainment vs Intent: What Whatnot Actually Sells 21:18 - The Supply Side: Turning Sellers Into Real Businesses 29:22 - Expanding Into Europe & the Seafood Distributor Story 32:47 - Operating at Scale: Staying on Top of What Breaks 39:34 - What's Next: Cars, More Countries & Better Businesses Resources: Follow Grant LaFontaine on X: https://x.com/GrantLaFontaine Follow David George on X: https://x.com/DavidGeorge83 Follow Whatnot on X: https://x.com/whatnot Stay Updated: If you enjoyed this episode, be sure to like, subscribe, and share with your friends! Find a16z on X: https://twitter.com/a16z Find a16z on LinkedIn: https://www.linkedin.com/company/a16z Listen to the a16z Show on Spotify: https://open.spotify.com/show/5bC65RDvs3oxnLyqqvkUYX Listen to the a16z Show on Apple Podcasts: https://podcasts.apple.com/us/podcast/a16z-podcast/id842818711 Follow our host: https://x.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see http://a16z.com/disclosures. [Source](https://www.youtube.com/watch?v=XqEr7hk89HY)

[1:06] Grant LaFontaine: 等一下，我玩宝可梦卡还能赚钱？我这么喜欢它，居然能靠它赚钱！ [9:02] Grant LaFontaine: 谁会想在 QVC 上买东西啊？那都是老奶奶才去的地方，跟我同龄的人绝不可能去 QVC 购物。 [9:02] Grant LaFontaine: 用户根本不在乎什么大盘市场（users don't give a **** about the market） [12:41] Grant LaFontaine: 在任意给定的一天中，Whatnot 上的绝大多数用户根本不会发生任何购买行为 [20:04] Grant LaFontaine: 好玩（Fun） [23:42] Grant LaFontaine: 把绝大部分价值留给商家 [27:07] Grant LaFontaine: 凡是现实中有人愿意为其开设线下实体零售店的商品，就一定能在 Whatnot 上运转起来。 [32:29] Grant LaFontaine: 度量一切（measure everything） [37:00] Grant LaFontaine: 人们来到 Whatnot，归根结底是为了人与人之间的真实连接

[12:41] Grant LaFontaine: Users don't give a ---- about the market. They care about whether they're going to enjoy the thing. [3:49] John: a horrific name [30:00] Grant LaFontaine: durable value medium [32:47] David George: relentlessness in execution [39:34] David George: maniacal, relentless focus on the seller experience

[0:00] If you look at e-commerce today, you [0:02] have to know exactly what you're looking [0:03] for. [0:04] >> Live commerce is over a third of all [0:06] commerce in Asia. Is that right? [0:08] >> 30 [music] to 40% of all commerce is [0:10] live commerce in China. [0:11] >> What is it in the US today? [0:13] >> Single digits really. The experience is [0:14] really broken. Transactions weren't [0:16] integrated well. Shipping wasn't [0:17] payments, discovery, CX, and we just [0:20] thought we could build a better one of [0:22] those. [0:22] >> I want to start all the way back when [0:24] you were a kid. Your first online sale [0:27] was a a [snorts] holographic Pokemon [0:30] card. [0:30] >> At [music] that point in time was when [0:32] like eBay was getting started and I was [0:33] like, "Wait, you can make money from [0:35] Pokemon cards? Like I love Pokémon cards [0:37] and I can make money from doing it." The [0:38] first one sold for 10 bucks. [music] [0:40] That's probably the true origin stories [0:42] of whatnot. [0:43] >> Users spend 95 minutes a day on the [0:45] platform. [0:46] >> Yeah, about 95 minutes. [0:47] >> It's incredible. [0:48] >> Users don't give a about the market. [0:50] They care about whether they're going to [0:51] enjoy the [music] thing. It provides [0:52] value and that's the experience that [0:54] makes it fun and keeps you coming back. [0:56] >> What are the most exciting things that [0:58] you're building [music] towards that you [1:00] want to grow into over the coming 12 [1:02] months? [1:02] >> Yeah. I mean, [1:06] okay, I want to start all the way back [1:08] when you were a kid. So, I understand [1:11] that your first online sale was a [1:15] holographic Pokemon card and that the [1:18] buyer mailed you a money order and you [1:22] cashed it at the post office and then [1:23] you shipped the card. So, this was like [1:26] a what age was this? This is like a [1:28] formative young age. [1:29] >> I must have been in third or fourth [1:31] grade I want to say. So, I don't How old [1:33] are you there? Like seven, eight, nine, [1:34] something like that. [1:35] >> Yeah. Okay. So there's there's all these [1:37] stories, you know, of a founder kind of [1:38] finding their market like this early. [1:40] Like tell me about like you kind of um [1:42] having that experience and making it a [1:44] formative piece of, you know, starting [1:45] whatnot eventually. [1:46] >> Well, I think like all kids back then, [1:48] like there was just a Pokemon craze. [1:51] Pokemon cards became big, you know, get [1:53] my dad to go bring me to the shop, [1:56] >> open my Pokemon cards. I had my binder, [1:58] had all my Charizards in it, had [2:01] everything. And I, you know, down this [2:03] down sort of this the street from where [2:04] I was living it was like a little like [2:06] lake and hangout area and my me and my [2:09] friends used to go there and we'd like [2:10] trade Pokemon cards and I'd try and [2:12] hustle people on like getting the better [2:14] cards. And then at that point in time [2:17] was when like eBay was getting started [2:19] and then their biggest competitor was [2:20] Yahoo Auctions. And [snorts] um I I [2:24] don't know how I found out about it, but [2:25] I found out about it. I was like, "Wait, [2:26] you can make money from Pokemon cards? [2:28] Like I love Pokemon cards. I can make [2:29] money from doing it." And so I [snorts] [2:31] got my dad to help me set up an account [2:32] and I you know posted my bunch of my [2:35] holographic cards online. The first one [2:37] sold for 10 bucks and it was the the [2:40] holographic um chancy card. [2:42] >> It was the Chansy card. Yeah. And the [2:44] the thing that always like um feels [2:46] funny about that is that at that point [2:48] in time the internet was so nent you [2:51] weren't doing payments online which is [2:52] just crazy. [2:53] >> Yeah. Yeah. This is like clunky. I mean [2:55] yeah early days like you know there was [2:57] no trust to do it. The infrastructure [2:58] didn't exist but there was no trust [3:00] either obviously. So [3:01] >> yeah and it's but it's also pretty wild [3:03] that like there was enough trust that [3:05] someone would send you [3:07] >> the like money order in the mail and [3:09] sort of like hope you would then send [3:11] the card back. So, I I remember I got [3:13] the money order like, "Okay, I can send [3:14] the card now. Dad, can you bring me to [3:15] the post office? I'm gonna wrap it up [3:17] and and um you know, send it off and uh [3:21] and so that is that's probably the true [3:23] origin stories of whatnot. Whatnot's [3:25] pretty funny in that uh almost all of my [3:28] experience in like a life like the big [3:31] the big bits, all of my like core jobs [3:33] have come full circle and whatnot." [3:35] >> Yeah. like Pokemon cards and I worked at [3:37] YouTube for a long time and so did video [3:40] and then I worked um I was I was working [3:43] at Facebook and I was working on [3:44] marketplace stuff and so I did some of [3:46] that and then whatnot. Sort of the [3:47] amalgamation of all those things. [3:49] >> Yeah, the perfect combination of all [3:50] those trends. Yeah, it's awesome. So, um [3:52] there's different stories about how you [3:54] and Logan decided to start the company. [3:56] How did you guys settle on doing this? [3:59] >> Yeah, so there's Logan's story and [4:00] there's my story. [4:01] >> Yeah. Okay. And and look, both of both [4:03] of them are like actually correct. I I [4:05] think Logan's just more um nostalgic [4:09] about like Logan's story is actually [4:10] like I don't know, [clears throat] maybe [4:11] it's like slightly better. So I was [4:15] Logan's origin story start in like [4:17] January [4:19] 2019, [4:21] I want to say. We we did a trip to Tokyo [4:23] over the holiday over like New Year's [4:26] and we were drinking in a bar and we [4:29] both of us had started businesses before [4:31] and so like a couple drinks we just [4:33] started talking about oh you know what [4:35] could we do one day and then we started [4:37] talking about you know building some [4:38] marketplace business we bought like 30 [4:41] or 40 domains and then like the trip [4:44] ended we got back to the real world and [4:46] like you know sort of nothing happened [4:48] for six seven eight nine months [4:50] basically And so that that's all true. [4:53] Like we talked about it and like we said [4:54] we're gonna do it, but then there was [4:56] like this [4:56] >> that wasn't the launch moment. Yeah. [4:57] >> Yeah. There was this like huge [4:58] intervening period. And so [5:00] >> um you know mine really was sometime [5:03] around like um you know October of 2020 [5:07] or 2019. Yeah. Uh, and um, you know, I [5:12] was like, Logan, remember we talked [5:13] about doing those businesses and we [5:15] brought those hundred domains when we [5:16] were drinking in a bar in [laughter] [5:18] Tokyo that we never that we never used? [5:21] Uh, maybe maybe now is the time. Logan [5:23] was like wrapping up his job. And um, [5:26] you know, I still had mine, but I was [5:28] like, you know, I need to go back to to [5:30] building some stuff. And we could see a [5:33] bunch of our friends getting back [5:35] neglecting. So it was like Funko Pops [5:37] and comic books and sports cards and I [5:39] was like you know I know a little bit [5:40] about that that industry and you know [5:44] I'd always dabbled even after you know [5:46] Pokemon cards I did um sneak I got into [5:49] sneakers for a while and then I was [5:51] doing some of the Funko stuff for a [5:53] while and so I sort of like knew that [5:55] market and it just like just hasn't [5:57] changed in a really long time and like [5:58] everyone uh like us now is is is older. [6:02] They're starting to have some disposable [6:04] income. They grow up on a mobile phone. [6:06] I think there's something around [6:08] collectibles and social commerce. Like, [6:11] you know, if someone's going to figure [6:12] it out, maybe why not us? Let's see what [6:14] we can do. [6:14] >> Yeah. [6:15] >> Uh, and so then December 2019, we both [6:18] quit our jobs and then we went full-time [6:20] and launched our authenticated Funko Pop [6:24] marketplace. [6:25] So, I got to ask, was not one of the [6:27] domains that you had bought? [6:29] >> Uh, no. That would have been Can you [6:31] imagine? Um, [laughter] [6:33] pretty good. That would have been [6:34] amazing forite. Yeah, it would have been [6:35] amazing actually. So the the the origins [6:37] of whatnot name which everyone always [6:40] asks cuz like whatnot's like a very very [6:42] memorable good name. We've had it for a [6:43] very long time. [6:45] >> Um so actually one of the domains we [6:47] bought again think we were drunk in a [6:48] bar in Tokyo so it wasn't as if our [6:50] judgment was [laughter] immensely good [6:53] was like always legit. And I'm not [6:56] joking because that was the first domain [6:58] for whatnot. And the reason it was the [7:00] first domain for whatnot is we're always [7:02] just obsessed with like let's just make [7:04] progress. Yes. [7:05] >> Let's just do a thing. I was like we [7:07] don't even have a product. We don't have [7:08] anything. I don't want to sit here and [7:09] talk about a branding exercise. [7:12] >> Yeah. [7:12] >> For days. [7:13] >> Just go. [7:14] >> Let's just go. [7:14] >> Yeah. [7:15] >> And so it was always legit. And then our [7:17] the first designer on whatnot, this guy [7:19] John, who I worked with on my my prior [7:21] startup before we were about to launch, [7:23] he's like, I cannot be associated with a [7:26] company called Always Legit. That is a [7:28] horrific name. And it was I named it and [7:31] it was a horrific name. But I was like, [7:32] it doesn't matter. We don't have a [7:33] product. Who cares what it's called if [7:35] you got nothing? [7:35] >> Yeah. [7:36] >> Um and and so he made us uh get into a [7:39] room, do a brainstorming exercise for [7:41] like an hour. We we got to like a list [7:44] of 10 names. We're like, "All right, [7:45] we'd be comfortable if we got any of [7:46] these. Let's go see if we can buy the [7:49] domain." And so I got like three, John [7:52] got three, and Logan got like three or [7:54] four. And um Logan got whatnot. And of [8:00] all those domains, all of them, you [8:02] know, would have cost like $100,000 plus [8:03] dollars and we had raised no money. [8:06] Logan got a hold of the guy who had [8:07] whatnot. And he, the guy goes, I'll sell [8:11] it to you for a Bitcoin. So I was an [8:13] early Bitcoin investor, so I had [8:15] Bitcoin. [laughter] [8:17] >> This is amazing. [8:18] >> And Bitcoin back then was worth $10,000. [8:20] So not like what it's worth now. And [8:23] Logan haggled them down to half a [8:25] bitcoin. [8:26] >> No way. Yeah. For five grand. [8:28] >> Yeah. For five grand. And so [8:30] >> I hope you guys held on to it. [8:32] [laughter] [8:33] >> I hope so too. What a I mean a great [8:35] price for that domain. And um so so then [8:39] the the plas service we were using it [8:41] was kind of funky because they hadn't [8:43] really used Bitcoin before. So anyways, [8:45] we sort of like sent the Bitcoin off and [8:47] then he sent the domain to us and he [8:49] sent us the wrong domain. He sent us why [8:51] not instead of whatnot. But we're like, [8:53] you know, not a bad also a pretty good [8:55] option, pretty good name. Um, but we're [8:58] like, let's try and see if we can get [8:59] whatnot because we do like we always had [9:00] this idea that we're going to sell lots [9:02] of different categories of items and [9:03] collectibles and so whatnot always felt [9:05] a little better. [9:06] >> Um, and luckily, you know, the I think [9:09] it took us 36 hours to get a hold of the [9:10] guy, but you know, 36 hours later, we [9:12] were able to to to get whatnot and we [9:15] were almost why not for a period of [9:16] time, which also would have been okay. [9:18] [laughter] [9:18] >> That's great. That's an awesome story. I [9:20] haven't heard that piece of it before. [9:21] Um okay so at the time that you guys [9:24] were getting started in the US Facebook [9:27] marketplace was working right in the [9:30] sense that there was a lot of activity [9:31] right and you could sort of see the [9:32] network effect behavior at that time is [9:34] that right and then [9:35] >> in China this had actually started to [9:38] happen right the live shopping trend had [9:40] become a category was early and then in [9:43] the US and the western world you had [9:46] like QVC so like talk about just the [9:48] state of the market when you guys were [9:51] actually starting and what you saw that [9:52] made you have conviction that this was [9:54] going to be a really good idea. [9:55] >> So I guess the the state of the market [9:58] um when we started it. So keep in mind [10:00] this is sort of December 2019 going in [10:03] just before co um so so one I'll preface [10:06] it with saying we probably weren't very [10:08] sophisticated on the state of the [10:09] market. Um we had no idea like live [10:12] shopping in Asia. We had heard of QVC. [10:15] >> Yeah. [10:16] >> Um and my impressions of QVC was like no [10:18] one wants to shop on QVC. like it's like [10:21] crannies shopping on QVC so no one my [10:23] age wants to shop on QVC [10:25] >> um and look there's a bunch of other big [10:28] marketplaces around um [10:30] >> but we sort of you know there was a [10:33] bunch for our market initially which was [10:35] collectibles uh you know there were a [10:36] bunch of players who were doing it but [10:38] sort of felt like it was a little bit [10:39] stagnant like the space hadn't changed [10:40] in a while so that was that was sort of [10:42] like the state of the market [10:42] >> and and by the way I would just say I [10:44] think it's probably an asset not a [10:46] liability that you guys didn't know much [10:49] about the market because if you knew a [10:50] lot about the market, you'd probably try [10:52] to build what existed in the market but [10:54] like iterate on it as opposed to a new [10:57] an entirely new thing. [10:58] >> Yeah. The I was 1,000% an advantage. [11:02] Yes. [11:02] >> Um because look, I I think the the [11:04] punchline and I can get into it is um [11:08] users don't give a about the [11:09] market. [11:10] >> Yeah. [11:10] >> Um they don't care about the market at [11:12] all. They care about whether they're [11:14] going to enjoy the thing. It provides [11:16] value. [11:16] >> Yeah. And so there there's this um [11:19] counterintuitive thing which is if you [11:20] built try and build a market it [11:23] oftentimes distracts from like are you [11:24] building a really great user experience. [11:26] Yeah. Because again like no one's the [11:28] consumer is not making a going like oh [11:30] this is a live shopping market and you [11:32] know I can transact all these things in [11:33] this market and they're going like well [11:34] no what are what are you going to [11:35] provide for me? What's the value? And so [11:37] um there were a bunch of players who [11:40] were trying to do that because they sort [11:41] of saw what was happening in Asian. It's [11:42] a it's a monster market. It's like a [11:44] trillion dollar plus market today. Um, [11:46] and we were like, we don't know. Uh, we [11:47] were building our Funko Pop marketplace [11:49] and we were talking to our customers [11:50] every single day and then we saw our [11:52] customers hacking social media in [11:54] interesting ways and the way they was [11:55] were trying to hack it by was selling [11:57] like um the selling items and live video [12:01] but the experience was really broken. [12:02] >> Yeah. [12:02] >> Uh, transactions weren't integrated [12:04] well. Shipping wasn't payments discovery [12:07] CX. Um, and we just thought we could [12:11] build a better one of those and it would [12:12] solve uh the problems of the customers [12:15] we had in the platform. Um, Logan locked [12:18] himself in his room for a few weeks, [12:20] came out with the first version of the [12:21] product and um, I went live, sold out, [12:25] sold a lot of items and the rest is sort [12:26] of, you know, history from there. [12:28] >> That's incredible. Yeah. Awesome. The [12:31] So, it's interesting to think about what [12:33] you all provide. So, I think it's a [12:34] super insightful way to say like [12:37] customers don't give a about a [12:38] market. Like they want a like they want [12:40] a product that's compelling to them. [12:41] Like what do you what value do you [12:42] provide to them? [12:44] Shopping often starts with intent. So, [12:48] how much of the idea of this is like [12:51] entertainment versus intent and like how [12:54] do you think about the distribution of [12:56] the behavior of the buyers and what [12:57] they're getting out of you? Yeah, I [12:58] think um [clears throat] whatnot is sort [13:00] of like analogous to visiting a store. [13:04] >> Yeah. [13:04] >> Um which is I don't know if I go to the [13:07] mall or I go to set of retail shops. [13:10] >> I'm sort of like I I kind of know I want [13:12] a thing but I don't know exactly what [13:14] that thing is. I'm going to go walk in [13:16] that shop. I'm going to go browse [13:17] around. And so [13:19] >> I don't what have I what have I bought [13:20] recent? I bought some some new shoes. [13:22] And so I I went on to go get some new [13:24] sneakers. I was like, "Okay, I need some [13:26] sneakers. I'm going to type in the [13:27] website and I'm going to go like I spent [13:29] 30 minutes browsing. Didn't know what I [13:31] wanted. I came away on the other side. [13:32] And that's sort of the whatnot journey. [13:34] Whereas if you look at a lot of the [13:36] e-commerce um companies that have been [13:38] built in the past, you know, prior to [13:41] us, you sort of know exactly what you [13:43] want. It's not really great for [13:44] discovery. It's not really great for [13:46] experiential or having fun or being [13:48] entertained. Uh so I think that's that's [13:50] sort of the the analog with whatnot. [13:52] Maybe you know you want a thing, but you [13:54] don't know what that exact thing is. [13:55] you're going to come on whatnot. Um, [13:56] you're going to look for it. You're [13:57] going to have some fun along the way. [13:59] >> Yeah, that's awesome. So, there's all of [14:01] those elements embedded in the product. [14:02] So, watching, chatting, bidding, [14:06] buying, returning. Are there pieces that [14:10] I'm missing? [14:11] >> Um, you mean just generally discovery? [14:14] >> Yeah, I think Yeah, I think you've [14:15] covered a lot. I mean the the the [14:18] one of one of the reasons [14:20] why whatnot has worked so well to to [14:24] date is that there's a lot of value [14:26] because it's such a you can do so many [14:28] different things. Sometimes you're going [14:29] to look for a great deal. Um a lot of [14:32] times you're going to just have an [14:33] experience and be entertained. So on any [14:35] [snorts] given day most people on [14:36] whatnot don't purchase anything. The [14:38] vast majority of people don't purchase [14:39] anything on whatnot. They're just going [14:41] and they're watching and they're talking [14:42] with people. Yeah. [14:43] >> Yeah. Um, and so that's that's why I I [14:46] continue like it's not a perfect analog, [14:47] but I I like the sort of analog of like [14:49] going to the shopping mall. Like you'd [14:51] go to the shopping mall, I hang out with [14:52] your friend, at least, you know, when I [14:53] was growing up, hang out with my [14:54] friends, walk around the mall, have some [14:56] fun. You know, many times I wouldn't [14:58] purchase something, just go there for an [15:00] experience. I am struck by obviously [15:03] we're longtime investors with you and so [15:05] you know this is this is not new news to [15:07] us but it is striking when I describe it [15:10] to other people how big the market for [15:12] this is and how big whatnot actually is [15:15] so you know you talked about Asia live [15:17] commerce I think the latest is it's over [15:20] a third of all commerce in Asia is that [15:23] right [15:24] >> yeah I've heard anywhere about 30 to 40% [15:26] of all commerce is is live commerce in [15:28] in China [15:29] >> yeah and then what is what is it in the [15:31] US today? [15:31] >> Oh, I I mean sing single digits. Really? [15:34] >> Single digits. And do you think there's [15:35] any is there any reason why you don't [15:37] think that that gap could close over [15:39] time? [15:40] >> No, I think it's it's clearly going to [15:41] close. Um it's just a richer way in [15:45] which to to sell and build a business. [15:48] Um and it's it's just more efficient. [15:50] So, you know, um if you're a [15:55] fashion brand or store, you could go and [15:59] set up a brick-andmortar shop on, I [16:01] don't know, pick a famous streetth Fifth [16:03] Avenue that cost you millions of dollars [16:05] and and rent [16:06] >> and you're restricted to a set of people [16:08] who come to your shop at a, you know, 9 [16:10] to5 on [16:12] >> um whoever is in New York City. [snorts] [16:14] Um and like that's a great experience. [16:16] You can feel it. You can talk to people. [16:17] You can discover things one whatnot. you [16:19] can set up for $0 and you can have an [16:21] entirely global audience that is, you [16:23] know, unbounded. [16:24] >> Yeah. [16:24] >> And so I think that that value prop [16:26] means if you're a business um you know [16:30] selling consumer goods, uh I think [16:33] you're eventually going to get into live [16:35] commerce. I don't think you're going to [16:36] be able to ignore it. I think you're [16:37] going to be able to bring in lots of [16:38] audience and that's why you see that 30 [16:40] or 40% number uh in China. [16:42] >> Yeah. So it's like the if you think [16:44] about just sort of marketplace dynamics [16:45] and the network effect the supply side [16:48] is so compelling because it's zero it's [16:50] effectively zero cost and with a very [16:53] compelling supply uh dynamic lots of [16:57] sellers that that drives demand right [17:00] >> yeah it makes total sense I mean the [17:02] numbers are just staggering you said [17:03] it's about a trillion dollars in in Asia [17:06] um that's a huge number I looked it up [17:08] before this so uh I think US e-commerce [17:12] is about a trillion too if I have that [17:16] roughly right. Um, and one of the [17:18] historical knocks I've heard people say [17:20] about whatnot is well, you know, Etsy is [17:23] only so big or eBay is only so big. Uh, [17:27] and I always find it funny when people [17:29] mention that, you know, e eBay was [17:31] actually, if you go back something like [17:33] 10 years, it was like 15% of the overall [17:36] market. And it's 80 billion of volume [17:38] still, by the way. it was like 15% of [17:41] the overall market. Now it's less than [17:42] 1% of the market. So the big change [17:45] that's happened is the market has [17:46] actually just grown beyond it. And so I [17:50] always say it's way too limiting to [17:51] think about, you know, what your market [17:53] is or what you can go get to think about [17:55] that. e eBay is sort of a a proof point [17:57] to me that there's at least an $80 [17:59] billion market of longtail commerce to [18:01] go get online. [18:04] But the real thing is like how much of [18:06] the gap between 80 billion and 1.2 two [18:08] trillion do you go capture because the [18:11] experience is so much better. Is that a [18:13] would you agree or disagree with all [18:14] that? [18:15] >> Uh yeah, look I I think um [18:17] [clears throat] [18:18] you know to tear it apart a little bit [18:21] um people will peg us to eBay because we [18:24] sort of got started in collectibles and [18:25] that's where you know eBay got started [18:27] but actually the live format is just [18:29] very very different. [18:30] >> Yes. [18:31] >> Um it means that a whole swath of [18:34] different types of sellers can use the [18:36] product. [18:37] >> Yes. anywhere from, you know, when eBay [18:41] got started, a lot of it was like [18:43] smaller resellers. If you look at [18:45] whatnot today, it's fundamentally like [18:47] small business and then we're getting [18:48] into brands. So, and a lot of those [18:51] brands and small businesses would [18:53] typically not be on eBay. [18:54] >> Yeah. They would never Yeah. wouldn't it [18:55] wouldn't make sense. Yeah. Maybe they [18:57] have a brick-and-mortar shop or they [18:59] have a Shopify store or, you know, a [19:01] bunch of things like that. So, I think [19:03] the the seller base is inherently much [19:05] more diverse. And then the thing with [19:08] live is it's it's um demand expansionary [19:13] because um you go in and discover new [19:17] things you didn't know that you would be [19:20] interested in. [19:21] >> Yes. [19:21] >> Uh and so if you look at e-commerce [19:24] today, it's so you have to know exactly [19:27] what you're looking for. So all it's [19:29] really done is take a bunch of offline [19:31] sales and bring them online in a more [19:34] efficient way. But you still got to know [19:36] exactly what you want. And as like live [19:38] commerce [19:39] >> uh proliferates, it's just going to [19:42] expand that market. So, not only could [19:43] it be 30 or 40% of that market, [19:46] um that market is going to get a lot [19:48] bigger. [19:48] >> Yeah. It's sort of like it would be more [19:50] akin to not so much like e-commerce, but [19:52] more like the inception of department [19:56] stores or shopping malls. Yes. [19:57] >> Right. Which is like actually expanded [19:59] the the demand characteristics of the [20:01] market. [20:01] >> Yeah. Exactly. [20:02] >> Yeah. That makes total sense. So the [20:04] other thing that is so striking about [20:06] the demand side and then I want to talk [20:08] about the supply side in great detail [20:10] because I think that's one of the most [20:11] exciting things about the platform is on [20:14] the demand side I understand users spend [20:16] 95 minutes a day on the platform. Is [20:19] that right? [20:19] >> Yeah about 95 minutes. [20:20] >> It's incredible. [20:21] >> That's like a shocking number and puts [20:24] you on par with social platforms like [20:27] the entertainment platforms. [20:28] >> Yeah. And it's because it's fun. Yeah. [20:29] Yeah, you know, if you were if you were [20:31] to sort of simplify whatnot down on like [20:33] the buyer value prop is it's a fun [20:34] experience to shop and there's not a [20:37] whole lot of fun experiences shopping [20:39] online. Um, and you know, historically [20:41] you'd go to a shopping mall and you [20:44] know, the dynamics there are all [20:45] changing so it's not even as fun to to [20:47] do that anymore. And and as I said [20:49] earlier, most people aren't even [20:51] purchasing. They're actually just going [20:52] to have a good time and experience it. [20:54] Most people don't p like you know [20:57] greater than 80% of people don't [20:58] purchase on a given day. They're just [20:59] there watching, having fun. [21:01] >> It's amazing. Yeah, it's an it's an [21:03] incredible sort of blending if you go [21:05] back to your history of entertainment [21:08] plus filling the shopping need um on the [21:11] demand side with new discovery and over [21:14] 80% of people not actually even [21:16] transacting. It's like clearly a form of [21:17] entertainment. I actually I spend a lot [21:19] of time on there uh entertaining. I [21:21] transact too, but uh yeah, I I get it. [21:24] That makes sense. Um, okay. So, that's [21:26] like the origin story, like the layout [21:29] of the market at a high level. And I [21:30] want to go into categories and stuff, [21:32] but I'd love to talk about the supply [21:34] side, like the sellers on the platform. [21:38] Talk about the scale that you have of [21:41] sellers. Um, and you know, this is sort [21:44] of one of the underrated things. By [21:46] being a marketplace, you enable all this [21:48] new business creation or business growth [21:50] of existing businesses that otherwise [21:52] would not exist. So I'd love to explore [21:54] that a little bit. [21:55] >> Yeah. I mean I think the one of the most [21:57] fun things about whatnot is it's it's [21:58] fundamentally empowering small business. [22:00] And if you look at a lot of what's [22:02] happened again in e-commerce over the [22:04] past 20 or 30 years is it it got so [22:09] efficient and one of the ways to make it [22:10] so efficient is to obscure who you're [22:12] buying from and obscure the ability for [22:15] people to like build brands and [22:16] businesses around it. And whatnot is [22:18] sort of the polar opposite. if you're on [22:20] whatnot, you're going to have your [22:21] store, you're going to have your brand, [22:22] you're going to have your customers, [22:24] they're going to come back to you. And [22:26] so, you know, I think that's one of the [22:27] one of the best bits of the platform is, [22:29] you know, we're generating billions and [22:30] billions and billions of dollars for [22:32] small businesses. And these are anywhere [22:34] from, you know, one and two person [22:36] operations all the way up to maybe a [22:38] couple hundred folks at the high end. [22:39] >> That's amazing. Yeah. The uh I I think [22:42] you've compared it in the past to like a [22:44] local shop owner or like a local pub [22:46] owner. And if you think about what [22:49] whatnot allows that person to do [22:51] compared to say Amazon which is like [22:54] highly utilitarian [22:56] um you know highly gamified and like [22:59] competition among first party third [23:00] party international sellers like it's [23:02] just an entirely different experience on [23:04] the sell side [23:05] >> um that will allow someone who's [23:06] creative and ambitious to thrive maybe [23:09] in a way that they wouldn't otherwise on [23:11] those platforms. Is that fair? Yeah, I [23:12] think it just empowers small business in [23:14] a way where maybe the the prior trend [23:17] was to almost disempower small business [23:19] and obscure them away. [23:22] Yeah. Yeah. That's well said. That's [23:24] well said. Um so you have really [23:28] interesting insight into all of these [23:30] different types of entrepreneurs, right? [23:32] So, um I'd love to maybe start with [23:34] just, you know, the original categories [23:36] like collectibles, cards, like what are [23:38] you seeing uh about the behavior in that [23:41] market? [23:42] >> Um [clears throat] so I think we're [23:45] we're seeing a continuation of the [23:46] trends that we saw when we first started [23:48] the business, which was one of one of [23:50] the reasons we we we came back into [23:52] collectibles. We saw a bunch of people [23:53] getting back into the market. it was [23:55] people like us who were had maybe some [23:57] money and were buying things that were [23:59] nostalgic from childhood and getting [24:00] back into unique um you know things that [24:03] they loved. Um and that trends basically [24:06] continued unabated for the past 6 and a [24:09] half years. And so um you know I think [24:13] collectibles is is largely going like [24:16] mainstream now for like lack of a better [24:18] term. You have you have mainstream [24:20] celebrities in it. You have athletes in [24:22] it. And so I think a lot of people [24:24] thought collectibles thought think of [24:26] collectibles like a niche thing, but [24:28] it's actually becoming a core part of of [24:29] pop culture. And I think over the past [24:32] six years, we've just seen that [24:33] evolution continue to grow and grow and [24:35] grow and grow. And you know, in fact, [24:37] you know, a lot of the new art that [24:39] people are collecting is now like [24:41] trading cards with that are going for [24:43] millions of dollars. And you know maybe [24:46] your parents if they were doing high-end [24:48] art or you know you were lucky enough to [24:50] have parents who could do that they're [24:51] buying you know famous artists and now [24:53] people are buying like the one of one [24:56] you know LeBron James refractor or [24:59] something like that. [25:00] >> Yeah. Yeah. Yeah. Yeah. It's super cool. [25:01] Uh yeah and it's obviously there's an [25:03] entertainment element of that as well. I [25:05] I I've been struck by the scale of some [25:07] of the businesses that are on whatnot. [25:08] So obviously you're empowering I think [25:11] millions of small businesses but you [25:13] know if done right uh a business owner [25:16] can actually build something north of [25:17] $100 million of revenue and highly [25:20] profitable all all on the back of [25:21] whatnot. [25:22] >> Yeah. You know the biggest businesses on [25:23] whatnot are probably doing upwards of [25:25] hund00 million in in revenue and you [25:27] know with like very strong ebida margins [25:30] 20 30 40 plus percent. you know what we [25:34] what we try and do um you know so if you [25:36] looked at whatnot last year we did you [25:38] know over eight billion in sales and um [25:40] so far this year we've done over eight [25:41] billion in sales and we don't we take a [25:44] small fraction of that and you know we [25:46] we and we'd like to we're we want to [25:48] we're not raising fees or doing any of [25:50] that. It's really important that the [25:51] businesses take the majority of that and [25:53] then our business grows with with all [25:54] those businesses. So [snorts] we just [25:56] spend you know all of our time trying to [25:58] help build great businesses and make [26:00] sure that people who are transacting [26:01] with those businesses have a great [26:02] experience. Yeah. Yeah. It's awesome. [26:04] It's it's it's one of the great and [26:06] probably lesser known, less covered [26:09] stories of small business empowerment in [26:11] the US and increasingly now in Europe, [26:13] which we can talk about, but it's uh [26:16] >> it's super exciting. Um I I want to [26:18] shift gears now. You know, that's [26:20] obviously a little bit on the supply [26:21] side. Um you know, talk about some of [26:23] the categories that you're working on. [26:25] Obviously, you know, started with [26:27] collectibles and and cards and then now [26:29] you have success, you've had a lot of [26:31] success in fashion. um which is [26:33] obviously one of the biggest categories [26:34] in commerce and there's many others. So [26:35] just talk about the category expansion [26:38] evolution over time. [26:39] >> Yeah, I mean I guess [clears throat] if [26:40] you look at whatnot today, we're selling [26:42] things in 100 plus categories. Um and [26:46] then we're [clears throat] always [26:47] investing in new areas to help people [26:48] grow their business. So some of the fun [26:50] ones today are like fresh food and drink [26:53] where you can buy fresh fish sourced [26:55] from the dock in San Diego or um [26:59] >> by the way Ben Ben and I started buying [27:00] it after dinner. [27:01] >> Yeah. Yeah. Yeah. Yeah, all that. It's [27:03] awesome. Um, another one of my [27:06] favorites, there's a a barbecue [27:07] restaurant where you can get amazing [27:09] barbecue and, you know, I have a bunch [27:12] of it sitting in my freezer. [laughter] [27:15] Um, you know, and I just I think, you [27:17] know, the fresh it's it's early days in [27:18] fresh food for us, but I'm excited to [27:20] buy it because I think the food supply [27:22] chain is so opaque in the US and it's [27:24] really nice to know where your food [27:25] comes from and be able to get the story [27:27] around it. um you know, flipping gears. [27:30] Uh I don't, for whatever reason, all my [27:32] friends golf now. I'm at that age in my [27:34] life. [27:34] >> Because you're getting old. Yeah. [27:35] >> Yeah. Um my hip's bad, so I can't do any [27:38] other real sports. Um and so golf on [27:41] whatnot is having a huge surge, you [27:43] know, that's growing hundreds of percent [27:45] year-over-year. We have all these golf [27:46] shops selling. And so that's, you know, [27:48] another another fun category that uh [27:51] I've been buying in and the team's been [27:52] investing in. And you sort of there's so [27:55] many beyond that even. [27:56] >> Yeah. It's a [clears throat] sort of [27:57] like anything that's highly considered, [28:00] right, should be a fit for whatnot [28:02] where, you know, an an explanation, a [28:06] personality that goes along with it, you [28:07] know, a story behind it, like would [28:09] would [28:10] >> I think anything that someone would [28:12] build a retail shop around would work on [28:14] whatnot. [28:15] >> Anything Anything that would be in the [28:16] shopping mall. Yeah. [28:17] >> Just literally anything because if you [28:18] were going if you were going to go visit [28:19] a store, well, you can do it from your [28:21] couch on your phone. [28:22] >> Yeah. That's awesome. What about uh [28:24] clothing and apparel in particular? [28:27] >> Um I mean women's fashions woman is now [28:29] our biggest category. [28:31] >> Wow. [28:31] >> Um and you know there's there's a huge [28:34] range of of stuff in there. Um I'd say [28:36] you know the dominant is is actually [28:38] sort of looks like almost TJ Maxx-L like [28:41] yeah which is like old season inventory [28:43] really good pricing and so you can get [28:44] some like amazing deals on like great [28:47] brands. Same with sneakers on whatnot. [28:50] Um, it's almost I forget the name. Like [28:52] I I when I was growing up there'd be [28:54] those like discount shoe stores where [28:56] you get like the nights for like 40 [28:57] bucks or something. There's there's a [28:58] lot of that and it's just like great [28:59] deal. [29:00] >> I shopped I shopped at a lot of those. [29:02] >> Yeah, it's like if you know you there's [29:04] there's nothing that beats a good deal [29:05] when you're shopping. [29:06] >> DSW. [29:07] >> Yeah, D. That's that's the one. It's [29:09] like the TJ Maxx shoes. Yeah. [29:11] >> Yeah. And so there's there's a a ton of [29:12] that popping up in like the fashion [29:14] apparel side and and Live's really [29:16] great. helps you move a lot of volume [29:18] and that when you can move that much [29:19] volume you're willing to take the [29:21] discounts and it gives customers a [29:22] really good price. [29:23] >> Yeah, 100% 100%. Um the other part of [29:27] you know expansion is obviously you've [29:29] had a tremendous amount of success in [29:31] the US. Um talk about the expansion into [29:33] Europe and what's worked, what's not, [29:36] what are the learnings. [29:37] >> Yeah. Um so today I think we're in about [29:39] we're in 10 countries. So, uh, US and [29:42] Canada and and North America, uh, five [29:44] countries in Europe, and we're also now [29:46] in Japan and Australia. So, we're we're [29:49] sort of, um, you know, have a good chunk [29:51] of the globe covered. Um the the I think [29:54] the most interesting thing that that I [29:57] don't that I think would [clears throat] [29:58] be counterintuitive to folks is that [30:00] we've just seen it work in the same way [30:03] across all of those countries, [30:05] >> which is there's um all the small [30:07] businesses that we've helped create in [30:09] the US, we've been able to help create [30:10] there and collectibles and fashion and [30:12] electronics and a lot of the dynamics [30:15] are incredibly incredibly [30:18] um incredibly incredibly similar. Uh and [30:22] so um you know we're just incredibly [30:24] bullish that like live is an incredibly [30:27] durable value medium in which to shop [30:30] from. [30:30] >> So one of my favorite stories is the is [30:32] the recent one that you've you told us [30:34] about the the seafood distributor uh who [30:36] messaged you on LinkedIn and uh first of [30:39] all like what an inbound. Uh and and [30:41] then you went live the following week [30:43] and he sold I think between three and [30:45] $4,000 of fish during the pilot phase. [30:49] Um, so what happened that caused that [30:52] and like how did that even come about? [30:55] >> Um, [30:57] again I I think the the punchline there [30:59] is that live is just a really good [31:00] median to discover things, right? And so [31:03] his story is, you know, he's been in the [31:05] fish business for a while. Deeply knows [31:07] what he's selling. He's going to bring [31:09] you the the freshest um fish from San [31:14] Diego at that time of year and he's [31:15] rotating the stuff in when it's [31:17] seasonal. Um, and he he's so incredibly [31:20] knowledgeable. So, you watch a show and [31:21] you see this type of fish you've never [31:23] heard of and you talk about how it's [31:25] sustainable and, uh, that it just got [31:27] off the pier today and here's a [31:29] fisherman who caught it. That's just a [31:31] great experience. And so, um, you know, [31:34] we can take him and we can push him to [31:37] whatnot now has a large, you know, buyer [31:38] base and and, you know, it's pretty easy [31:41] then to generate thousands of dollars an [31:44] hour in sales um, with that experience. [31:46] And so, [31:47] >> yeah. And imagine him trying to sell [31:49] over the internet at large. Like there [31:51] actually like is like he can access his [31:53] local market, but there's probably no [31:55] way for him to really access a [31:57] nationwide market for something like [31:59] this. [31:59] >> And I think it's really hard like buying [32:02] food on the internet like fresh fish [32:04] like cuz it's one of it's one of those [32:06] things you're like wow fresh fish on the [32:08] internet like how do I know if it's good [32:09] if I can get sick from it? [32:11] >> Um and so video just really enhances the [32:13] trust. [32:14] >> Yeah. Yeah. Yeah. and like a a real [32:15] identity of the person the other end you [32:17] can ask questions you can know how he's [32:19] going to pack it [32:20] >> and if you're on a static website like [32:23] well I hope it's good [32:24] >> yeah [laughter] [32:25] that's true that is very true okay so if [32:29] you think about the platform and you've [32:31] talked about the scale that you guys are [32:32] running it's just massive massive scale [32:36] um and so you one of the things I've [32:40] used to describe you and Logan is [32:43] relentlessness like you guys are [32:45] relentless in execution. [32:47] What are some of the things that you [32:49] encounter with that amount of scale that [32:52] you constantly have to be on top of? And [32:55] I think this is a big like obviously now [32:56] you're big and you have a network [32:58] effect, but I think this is one of the [33:00] big underrated things that allowed you [33:02] to win in the early days is that you [33:04] guys were constantly on top of anything [33:06] that could break. So I'd love to [33:08] understand some of those big things. [33:10] >> Yeah, I mean it's it's it's challenging. [33:13] Um I mean you know there's so many [33:15] dimensions to to trust. So there's sort [33:17] of like is the platform reliable and is [33:20] it working and is it stable and is it [33:22] fast and so um you know we measure just [33:25] everything. [33:26] >> Yeah. [33:26] >> Um like I I think the the punch line on [33:28] trust is measure everything. Um [33:33] and uh we have a invest a lot in it. So [33:37] I think um trust and safety particular [33:39] on whatnot is about 40% of employees. [33:42] Yeah. So, it's au it's a huge huge [33:43] investment from us because it's it's a [33:45] challenging one and you don't want to [33:46] put your [33:47] >> credit card in a place you you don't [33:49] trust. You certainly don't want to buy [33:50] fresh fish from a place you don't trust, [33:52] >> right? [33:53] >> Um and and so so so those you know I [33:57] think lot of measurement, lot of user [33:59] feedback um and then a big just people [34:04] investment to get the thing done is is [34:07] sort of what drives it. Um, and then you [34:10] know the problems are are pretty diverse [34:13] and you have to make sure you're [34:14] building the right set of things for it. [34:16] So as an example, [clears throat] we [34:18] have a a system in the background um [34:21] which is called our our rules engine and [34:23] it um [34:26] takes in billions of data points at this [34:28] point in sub [34:31] 1 second to be able to take down or find [34:36] any range of bad behavior on the [34:38] platform. So if you harass someone in a [34:41] live stream, the LLM will detect it. It [34:44] will get plugged into the our action [34:46] rules engine will look at the history of [34:48] the seller and based upon other signals [34:51] programmatically take them down, suspend [34:52] them, ban them, warn them. If you are [34:55] not shipping on time, the the rules [34:57] engine will do the same thing. If you [34:59] have a high refund rate, the rules [35:00] engine will do the same thing. And so, [35:03] um, and look, these are ever changing, [35:07] you know, problems and issues. [35:09] >> Waiting of scoring and things like that [35:10] will always change. Yeah. Um and and so [35:12] you constantly have to uh be on top of [35:15] these things. [35:16] >> Yeah. Yeah. I mean, you know, if you [35:18] think about just how many concurrent [35:20] users you have and how many transactions [35:22] you process per day. [35:25] >> Yeah. I I always the the the analog I [35:27] like for trust and safety is so you know [35:29] tens of millions of people use whatnot [35:31] every [35:31] >> Yeah. Yeah. [35:32] >> And so um the trust and safety team is [35:35] basically the police force and the the [35:37] legal system of of whatnot. You got to [35:39] write your policies. You got to be [35:40] people who enforce your policies. And [35:43] you know, in a city that has 10, 20 [35:47] million people, you're looking at a city [35:48] the size of Tokyo or I mean, maybe a [35:51] better analog would be like I mean, it's [35:53] a couple couple New York cities. [35:55] >> Yeah. Two New York cities. [35:56] >> Two New York cities. And um [35:59] uh you know, in in New York, um it's not [36:02] not perfectly run. Uh and so there's [36:05] going to be some bad things that happen. [36:07] and people get arrested. But at at our [36:09] scale, because of some of the systems we [36:10] build, there's actually far far fewer [36:12] Yeah. [36:13] >> problems. And and then of course, look, [36:14] if whatnot is a business for a small [36:16] small business, everything's on video. [36:19] So if you're so if you're a small [36:20] business, you want to build a great [36:21] business. [36:22] >> Any good business knows you got to have [36:23] repeat customers, you got to invest in a [36:25] customer experience. So our sellers [36:26] reinforce it. [36:27] >> Then you're on live video. [laughter] So [36:29] there's not, you know, you're not going [36:30] to get away with too much on live video. [36:32] You know, things things happen. Um, and [36:35] then you you sort of go, okay, we put [36:37] 40% of the company on, you know, law [36:39] enforcement effectively, their trust and [36:41] safety team. And so, um, it, you know, [36:44] it still means it's challenging because [36:45] if you're running the [36:46] >> the the um police and legislative system [36:50] of of New York City, that's a [36:51] challenging and hard job that you always [36:53] have to be on top of and be, you know, [36:55] looking for what the new areas are to [36:56] invest in and and improve. [36:58] >> Yeah. Yeah. Absolutely. This makes total [37:00] sense. um the so it's actually a good [37:03] segue um into the topic of AI because [37:06] obviously you're using cutting edge AI [37:09] to to do a lot of the trust and safety [37:11] work. Maybe taking a step back with AI [37:13] though um you know maybe front facing in [37:17] the product side you know not not the [37:19] back end. Um you know you have a very [37:22] inherently human platform where buyers [37:25] and sellers are engaging as humans. What [37:28] are the learnings about how to employ AI [37:31] most effectively while also preserving [37:34] that human connection? Yeah, I mean I [37:37] think the good thing is some of the [37:39] platforms self-reinforcing which is um [37:42] [snorts] [37:43] you know if someone goes over top thon I [37:45] don't know building an AI avatar people [37:47] aren't going to come back to them right [37:49] it's not it's not a business because the [37:51] core the core of whatnot is you visit a [37:55] shop you know the shop owner and then [37:58] you know the people who frequent that [37:59] shop [38:00] >> and that's the experience that makes it [38:02] fun and and keeps you coming back [38:04] [snorts] [38:04] >> and so Anyone who's good is not um [38:08] trying to rip out the human experience. [38:10] Anyone who's good is is going to enhance [38:13] is going to try and enhance that. And so [38:14] then where where AI comes in is well, [38:18] how do you help someone keep it human [38:20] but run their business more efficiently? [38:21] >> Yeah. [38:22] >> So whether that's like listing items, we [38:24] can make that much better. Um we, you [38:27] know, when you're setting up a live [38:28] stream for the discovery systems, often [38:30] time you need to add a lot of metadata. [38:32] Well, we can infer all that metadata [38:35] with LLMs. Um, we're investing in it [38:38] from an analytics perspective. So, let [38:40] we'll tell you where your business is [38:42] doing well, where it's not doing well, [38:43] where you can improve it. And so, you [38:46] know, honestly, in like a fairly [38:48] succinct way, what we what we generally [38:50] do is um look [clears throat] at where [38:52] we're deploying AI and where it's being [38:53] efficient for us and then try and expose [38:56] it to the sellers. [38:56] >> Yeah, that makes total sense. And if I [38:59] hear you right, that doesn't involve [39:01] like giving sellers avatar tools or [39:05] something like that to [39:06] >> No, I mean, look at [39:07] >> try to build scale or something like [39:09] that. [39:09] >> Look, at the end of the day, we'll we'll [39:12] put energy into what's going to be good [39:13] for the buyer and it's going to be good [39:15] for the seller. Um, but today we don't [39:18] see a need for uh avatars because that's [39:22] not why people come to what? [39:23] >> Yeah, they come for the human [39:24] connection. Yeah, totally agree. That [39:26] makes a ton of sense. We've covered a [39:28] ton of ground like we've talked about, [39:29] you know, the platform today, the origin [39:31] story, supply side, demand side, whatnot [39:34] is definitively one of the most [39:36] interesting and unique companies that [39:38] we've invested in. Um, and you know, [39:40] I've been very fortunate to be part of [39:41] your journey. In closing, I would love [39:44] to hear just what is next, like what are [39:46] the most exciting things that you're [39:47] building towards uh that you want to [39:50] grow into over the coming 12 months? [39:52] Yeah, I mean in in some ways it's it's [39:55] relatively simple, which is um there's [39:57] just an incredible amount of work to do [39:59] to help more and more businesses grow on [40:02] whatnot. And so over the next 12 months, [40:05] 24 months, 36 months, a lot of it is [40:08] just like making sure sellers can run [40:09] better businesses on whatnot, giving [40:11] them the tools to run better businesses, [40:14] opening up and making the experience [40:16] better for more categories on whatnot. [40:18] So, one of the ones that I've been dying [40:20] to do um is like I want to do cars on [40:24] whatnot because you should be able to [40:25] buy a car on whatnot. And how cool would [40:27] it be to be able to like see the full [40:28] experience, get the video behind it? [40:31] >> Oh, this is that's an awesome idea. [40:32] >> Yeah, it's going to be um it's going to [40:34] be amazing. It might be not good for um [40:37] uh my bank account. [laughter] Um and [40:39] I'm I'm obviously obligated to to try it [40:42] out. [40:43] You got to dog food it. Um, and so, so I [40:46] think there's a, you know, just a huge [40:47] amount of investment just continuing to [40:48] make sure make people help people build [40:51] better businesses. Uh, and then, you [40:53] know, in order to do that, we have to [40:55] have a really great buyer experience. [40:56] >> Yes. Because we in order to build a good [40:58] business, people have to come back. And [40:59] so, we're going to continue to make sure [41:01] we're investing in customer support, [41:03] trust, and safety, bringing down [41:05] shipping prices, improving the selection [41:08] and quality of the products on the [41:09] platform. And that's that's really it. [41:11] Now, maybe we'll launch in some more [41:13] countries. We'll do some cool stuff like [41:14] cars. You know, eventually we'd like to [41:16] get into liquor, beer, and wine. So, [41:17] there's a whole litany of things that we [41:19] we're we're excited to do, but [41:21] fundamentally it boils down to help more [41:23] people build businesses and make sure [41:24] that the people transacting those [41:26] businesses are having a great [41:27] experience. [41:28] >> That's an awesome way to close. Your [41:30] maniacal, relentless focus on the seller [41:33] experience obviously drives the buyers [41:36] and then making the buyer experience as [41:39] affordable and as good as possible is [41:41] clearly working. So, um, thanks for the [41:43] time, Grant. So fun to be with [41:45] >> you. [41:46] [music]