内容来源：https://www.wired.com/story/ai-lab-walmart-openai-shaking-up-agentic-shopping-deal/

内容总结：

沃尔玛AI购物实验遇挫，ChatGPT内嵌电商功能转型

自去年11月起，沃尔玛与OpenAI合作，在ChatGPT中测试了一项名为“即时结账”的AI购物功能，允许用户在不离开聊天界面的情况下直接购买约20万种商品。然而，这项被寄予厚望的尝试并未取得预期效果。沃尔玛负责设计与产品的高级副总裁丹尼尔·丹克尔向媒体透露，该功能的销售转化率“令人失望”——在聊天机器人内直接完成购买的用户比例，仅为需要跳转至沃尔玛网站购买的商品的三分之一。

实验结果表明，当前阶段完全依靠聊天机器人接管电商购物流程的设想仍不成熟。OpenAI原本希望通过此类合作，对通过ChatGPT达成的交易抽取佣金以增加收入。但“即时结账”功能在实际使用中存在明显短板：它强制用户对每件商品单独结账，这与消费者习惯将多种商品合并下单、统一配送的期望相悖。丹克尔指出，用户担心“每买一件商品就自动结账一次，最终会收到五个包裹”，而他们实际希望所有商品能一起送达。

基于此，双方迅速调整策略。下周起，沃尔玛自研的聊天机器人“Sparky”将直接内置于ChatGPT中运行，形成“聊天机器人中的聊天机器人”。下个月，类似功能也将登陆谷歌的Gemini聊天机器人。新模式的核心改进在于：用户首次在ChatGPT中登录Sparky后，其在沃尔玛官网或App中的购物车将与ChatGPT内的购物车同步，从而支持跨平台、多商品合并结算，旨在还原更真实的购物习惯。

尽管“即时结账”功能整体表现不佳，但部分品类如维生素、蛋白补充剂等健康产品销量领先。丹克尔分析，这与GLP-1减肥药使用者通过ChatGPT咨询后获知需增加营养摄入的建议有关。此外，汽车用品、美容产品、家居硬件等单价较高、免收运费的品类，占据了该功能订单量的一半以上。

沃尔玛积极布局ChatGPT生态，源于其带来的用户增长潜力。丹克尔表示，目前ChatGPT为沃尔玛吸引新客户的速度已是传统搜索引擎的两倍。Sparky由沃尔玛基于开源生成式AI模型并结合其数十年零售数据训练的专业模型开发，可根据问题类型调用不同模型以优化回答质量。

对于未来，丹克尔认为购物完全自动化的愿景“可能有些牵强”，因为许多消费者仍享受为家人、家庭选购商品的过程。沃尔玛的目标并非用AI取代人的决策，而是通过Sparky等工具，在更多场景中为用户提供伴随式辅助，同时确保购物体验的自主性与可控性。公司表示，只要用户体验良好，不会阻止其他AI代理在其平台购物，但会避免出现错误订单、账单纠纷或客服压力激增等问题。

目前，沃尔玛App用户中已有半数与Sparky进行过互动。数据显示，使用Sparky的用户客单价较其他用户高出约35%。沃尔玛承认Sparky目前存在响应速度慢、答案质量不稳定等问题，并计划今年重点提升其主动性、个性化程度及在多业务部门（如药房）的应用能力。

中文翻译：

自去年11月起，沃尔玛允许部分ChatGPT用户在聊天界面内直接下单购买限定商品。但沃尔玛执行副总裁向《连线》杂志独家透露，这项功能的销售表现令人失望。

这一结果表明，由聊天机器人和AI代理主导电商的未来尚需时日，甚至可能永远不会实现。去年，OpenAI曾押注通过收取ChatGPT内购佣金来提升收入，并与沃尔玛、Etsy等商家合作推出了名为"即时结账"的智能商务功能。

沃尔玛在聊天回复中直接提供了约20万种商品，消费者只需向OpenAI提供收货和支付信息即可在ChatGPT内完成订单。但对于电视等商品，消费者仍需跳转至沃尔玛网站按传统方式购买。负责沃尔玛设计与产品的丹尼尔·丹克尔透露，聊天机器人内直接销售商品的转化率仅为需要跳转购买商品的三分之一。简而言之，即时结账功能遭遇了滑铁卢。

丹克尔表示，OpenAI和沃尔玛本可以花费数年时间改进即时结账"不尽如人意"的消费体验，但他赞赏OpenAI选择快速转向沃尔玛长期青睐的新模式。下周起，沃尔玛聊天机器人Sparky将入驻ChatGPT——形成聊天机器人嵌套聊天机器人的架构。类似配置也将在下个月登陆谷歌Gemini聊天机器人。

丹克尔认为新模式解决了即时结账的最大痛点：强制用户逐件购买商品。"消费者担心每选一件商品就自动结账，最终收到五个包裹，而他们其实希望所有商品能合并发货。"丹克尔解释道，"即便购物车里已有其他商品，他们也不愿经历拆分结账的体验。"

在即时结账功能中，维生素和蛋白质补充剂成为畅销品类。丹克尔指出，初次接触GLP-1减肥药的用户向ChatGPT咨询时，会收到增加营养摄入的建议。其他表现较好的商品通常价格较高，可免收附加运费或小订单费。总体而言，汽车用品、美妆、家居管理、五金工具等品类占据了即时结账订单量的半数以上。

在新模式下，沃尔玛用户首次在ChatGPT中遇到Sparky时需登录账号。他们在沃尔玛网站、应用及ChatGPT内的购物车将实现同步，以期更真实反映消费习惯。丹克尔描述道："消费者可能今天在应用里加购花生酱，明天选锡纸，结账前最后一刻又在网站添置生日礼物。Sparky的核心理念是让沃尔玛商店随时伴您左右，而非制造割裂的体验。"

沃尔玛有充分理由优化ChatGPT内的购物体验。丹克尔透露，目前聊天机器人带来的新客户转化率是搜索引擎的两倍，这可能因为ChatGPT的高频用户并非沃尔玛典型客群。但该零售商的价格优势、商品种类和广泛覆盖范围，使其产品频繁出现在ChatGPT的回复中。

丹克尔说明Sparky由沃尔玛自主开发，其底层融合了开源生成式AI模型与基于数十年沃尔玛数据训练的零售专用模型。"我们发现不同模型擅长处理的问题类型不同，因此建立了智能路由系统。"他表示，"这套系统永远不会固守单一模式。"

考虑到新的集成需求，这款聊天机器人被特意设计得灵活可变。"它能微调界面风格，使其在不同环境中都显得自然协调。"丹克尔补充道。

据《信息报》本月初报道，沃尔玛的新体验是OpenAI战略转型的组成部分，该转型旨在将结账流程嵌入合作方应用内部，但报道未说明调整原因。丹克尔在本月摩根士丹利投资者会议上谈及此次转型，但未透露具体数据。

OpenAI发言人塔亚·克里斯蒂安森表示，公司将专注于优化产品研究功能，同时赋予商家更多结账控制权。"我们感谢合作伙伴与我们一起探索学习。"她补充道。

丹克尔透露，沃尔玛已将部分商品排除在即时结账功能外，因为他们意识到"单件结账模式在某些场景下存在弊端"。例如购买电视时，消费者通常需要HDMI线等配件。在官网上，沃尔玛可以引导用户购买套装以避免安装困扰，而通过Sparky，沃尔玛将能在聊天机器人中复现这种体验。

零售商当初对即时结账功能热情高涨，因为当时服务ChatGPT用户的唯一方式就是跳转至商家网站。沃尔玛相信Sparky能带来"更无缝"的体验——用户无需重复输入已保存在沃尔玛的支付和配送信息，就能持续对话并完善订单。

虽然Reddit上有自称沃尔玛员工的用户批评Sparky，社交媒体上也难觅其好评，但该公司数据显示半数沃尔玛应用用户曾使用该功能。当人们通常用应用搜索牛奶、香蕉等日常商品时，他们会向Sparky咨询特色商品或寻求复杂问题解决方案。沃尔玛美国CEO戴维·古吉纳近期透露，Sparky用户的单笔订单金额比其他消费者高出约35%。

丹克尔承认Sparky响应速度较慢，且时常生成质量欠佳的回复，可能导致部分消费者认为其不可靠。他表示今年的重点是训练Sparky变得更主动，深化对个体消费者的了解，并拓展至药房等更多沃尔玛部门提供服务。

尽管沃尔玛正在推广Sparky，但该公司并未——也不打算——阻止其他AI代理在其网站购物。相比之下，亚马逊近期获得临时法院禁令，禁止Perplexity的自动化技术伪装人类进行购物。丹克尔强调，只要体验良好，沃尔玛愿意支持消费者使用的任何工具，前提是不出现错误订单、天价账单或过度依赖客服的情况。

"我们不想规定每位顾客必须遵循固定路径，"他表示，"也不会因推测性或假设性问题阻断创新。"

当被问及多少消费者会信任AI购物时，丹克尔给出了推测："认为购物将完全自动化的想法可能有些牵强。人们确实会为购置衣物、布置家居、为孩子挑选商品而感到兴奋。"沃尔玛希望将控制权交还用户，只是如今在更多场景中让Sparky陪伴左右。

本文节选自威尔·奈特《AI实验室》时事通讯，过往内容可通过此处查阅。

英文来源：

Since November, Walmart has let some ChatGPT users order a limited selection of products without ever leaving OpenAI’s chatbot interface. Sales have been disappointing, a Walmart executive vice president exclusively tells WIRED.
The results suggest that a future where chatbots and AI agents take over ecommerce is still a way off, if it ever materializes. Last year, OpenAI made a bet that it could boost revenue by charging a commission on purchases made through ChatGPT. It partnered with Walmart, Etsy, and other shops on an “agentic commerce” feature called Instant Checkout.
Walmart has made about 200,000 products available directly in chat responses, allowing consumers to provide their shipping and payment details to OpenAI and place their order within ChatGPT. For products like TVs, shoppers still have to open Walmart’s website to make a purchase the old-fashioned online way. Conversion rates—the percentage of users following through with a purchase of an item shown to them by ChatGPT—have been three times lower for the selection sold directly inside the chatbot than those that require clicking out, according to Daniel Danker, who oversees design and product for Walmart. Put simply, Instant Checkout has been a flop.
OpenAI and Walmart could have spent years trying to fix the ”unsatisfying” consumer experience of Instant Checkout, Danker says. He credits OpenAI for choosing instead to quickly move to a new system long favored by Walmart. Next week, Walmart’s chatbot, Sparky, will begin operating within ChatGPT—essentially a chatbot inside a chatbot. A similar setup will arrive in Google’s Gemini chatbot next month.
The approach solves what Danker says he believes is the biggest problem with Instant Checkout: It forces people to buy items individually. “They fear that when checkout happens automatically after every single item that they're going to receive five boxes when they actually just want it all in one,” Danker says. “They generally don't want to split the checkout experience, where it buys the one item, even though they had other items in their Walmart cart already.”
Among the items that have been available in the Instant Checkout experience, top sellers include vitamin and protein supplements. People new to GLP-1 weight-loss drugs asking ChatGPT what they need to know about them are receiving advice to increase nutrient intake, Danker says. Other items that have done well tended to be pricey enough that the orders weren’t subject to added shipping or small-basket fees. All told, the automotive, beauty, home management, hardware, and tools categories account for over half of Instant Checkout orders.
In the new experience, Walmart users log into Sparky the first time they encounter it in ChatGPT. Their basket from Walmart’s website or app and within ChatGPT will sync with another in the hopes of better reflecting people’s actual shopping habits. Consumers add peanut butter one day on the Walmart app, foil the next, and a birthday gift at the last second on the website before checking out. “When Sparky travels, it's the Walmart store meeting you where you are, instead of a completely broken experience,” Danker says.
Walmart has good reason to want to get the experience in ChatGPT correct. The chatbot is now bringing in about twice the rate of new customers as search engines, Danker says. He suspects that’s because the power users of ChatGPT are not typical Walmart customers. But the retailer’s price, selection, and broad geographic footprint mean that its products are showing up in many ChatGPT responses.
Sparky was developed by Walmart, Danker says. But it relies on open source generative AI models combined with some retail-specific ones trained on decades of Walmart data. “We're able to route certain questions to one model and certain questions to another because we find that the quality of answers differs,” Danker says. “It's never stuck in any one.”
The chatbot also is intentionally flexible, with the new integrations in mind. “It can take on slight tweaks to the look and feel, to make it feel like a natural part of other environments,” Danker says.
Shopping Shift
The new Walmart experience is part of a broader pivot for OpenAI to focus on having checkouts take place within embedded apps, the Information reported earlier this month, without providing a rationale for the change. Danker spoke about the shift at the Morgan Stanley investor conference this month but didn’t cite the data behind it.
OpenAI spokesperson Taya Christianson says the company wants to focus on improvements to help users research products, while giving merchants more control over checkout. “We appreciate our partners for learning with us,” she added.
Walmart has excluded some products from Instant Checkout because it knew “the single-item checkout experience is detrimental” in some cases, Danker says. For instance, when someone buys a TV, they likely need to buy accessories like HDMI cables. On its website, Walmart can nudge shoppers to buy a bundle to avoid a frustrating installation experience, Danker says. Through Sparky, Walmart will be able to replicate that in chatbots.
Retailers were eager to collaborate on Instant Checkout because the alternative at the time to serve ChatGPT users was by linking out to their websites. Walmart believes the Sparky experience will feel even “more seamless,” because users will be able to continue chatting and refining their order without needing to reenter their payment and delivery information already saved with Walmart.
Sparky has been criticized by people purporting to work for Walmart on Reddit, and testimonials for the chatbot are difficult to find on social media. But half of Walmart app users have engaged with it, according to the company. While people typically use the app to search for staples such as milk and bananas, they ask Sparky about exotic items or for solutions to more complicated problems. Walmart US CEO David Guggina recently said Sparky users spend about 35 percent more per order than other shoppers.
Danker acknowledges that Sparky is slow and generates weak responses often enough that some consumers might dismiss it as unreliable. Danker says the priority this year is training Sparky to be more proactive, getting it to learn more about individual shoppers, and making it helpful across more of Walmart’s many departments, such as the pharmacy.
While Walmart is pushing Sparky elsewhere, it hasn’t—and doesn’t plan—to block other AI agents from shopping on its website. Amazon, on the other hand, recently won a temporary court order barring Perplexity’s automated technology from masquerading as a human to make purchases. Danker says Walmart wants to support whatever tools customers are using as long as it’s a good experience. As in, there shouldn’t be erroneous orders, shocking bills, or an excessive need for customer service.
“We don't want to be prescriptive of the exact journey that every customer is going to take,” he says. “We don’t want to block things on a speculative or hypothetical concern.”
When it comes to how many consumers will trust AI with their shopping, Danker is prepared to speculate. “This idea that it will all become automated might be a little bit far-fetched,” he says. “People do get excited about shopping for clothes, for their home, for their children.” Walmart wants to leave users in control, just now with Sparky by their side in more places.
This is an edition of Will Knight’s AI Lab newsletter. Read previous newsletters here.