--- title: 你的AI客服學會的不是解決問題,而是讓用戶閉嘴 | daily.dev description: Klarna's 2023 announcement of replacing 700 customer service agents with AI became a cautionary tale by 2025, when the CEO admitted quality had suffered and... canonical: https://daily.dev/posts/ai--mj93nrhe9 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: 你的AI客服學會的不是解決問題,而是讓用戶閉嘴 | daily.dev og:description: Klarna's 2023 announcement of replacing 700 customer service agents with AI became a cautionary tale by 2025, when the CEO admitted quality had suffered and... og:url: https://daily.dev/posts/ai--mj93nrhe9 og:image: https://api.daily.dev/og/posts/MJ93Nrhe9.png og:image:alt: 你的AI客服學會的不是解決問題,而是讓用戶閉嘴 og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # 你的AI客服學會的不是解決問題,而是讓用戶閉嘴 **[Medium](https://daily.dev/sources/medium_js)** · 1 min read · 0 upvotes · 0 comments ## Summary Klarna's 2023 announcement of replacing 700 customer service agents with AI became a cautionary tale by 2025, when the CEO admitted quality had suffered and the company began rehiring human agents. The core problem: most companies measure AI customer service success using 'deflection rate' — how many conversations ended without human transfer — rather than whether customer problems were actually solved. Research shows self-service channels fully resolve only 14% of issues, yet companies report 60%+ deflection rates, meaning the gap represents customers giving up, not problems solved. Drawing on Nielsen Norman Group's five dimensions of AI chatbot quality (handoff willingness, flexibility, proactivity, emotional response, transparency), the piece illustrates how optimizing for deflection rate actively undermines each dimension. Real examples show chatbots blocking human transfers, failing to recognize their own product catalog, pretending to have capabilities they lack, and using scripted emotional responses. The silent consequence: 44% of customers who have bad AI service experiences simply stop buying rather than complaining, creating retention gaps that deflection metrics will never reveal. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What happened with Klarna's AI customer service after they replaced 700 human agents with AI in 2023? By 2025, Klarna reversed course and began rehiring human customer service staff after CEO Sebastian Siemiatkowski admitted the company overemphasized efficiency and cost cutting at the expense of service quality, calling that approach unsustainable. The company adopted a hybrid model where AI handles simple queries while humans take over complex or emotionally sensitive cases. _Anyone weighing AI versus human support tradeoffs can follow real-world case studies like this one on daily.dev._ ### Why is deflection rate a misleading metric for measuring AI customer service success? Deflection rate measures whether a conversation ended without being transferred to a human, not whether the customer's problem was actually solved. Analysis shows only around 14% of self-service interactions fully resolve issues, and even simple cases resolve just 36% of the time, yet companies commonly report deflection rates above 60%, meaning most of that gap reflects customers giving up rather than getting help. _Teams designing support KPIs can dig into resolution-versus-deflection tradeoffs like these on daily.dev._ ### What percentage of customers stop using a company after a bad AI customer service experience? Close to 44% of customers who lose trust in AI customer service simply stop purchasing from that company altogether, often without complaining or explaining why, leaving retention numbers to quietly decline. Separately, a Qualtrics survey of over twenty thousand consumers found nearly one in five AI customer service users reported no benefit at all from the experience. _Anyone tracking churn tied to AI support decisions can follow findings like these on daily.dev._ ## Similar posts on daily.dev - [Trends that are shaping the use of AI to improve CX](https://daily.dev/posts/trends-that-are-shaping-the-use-of-ai-to-improve-cx-i7myqwcfp) · TechCentral · 0 upvotes · 0 comments - [How AI Customer Support Agents Are Changing Service Operations](https://daily.dev/posts/how-ai-customer-support-agents-are-changing-service-operations-vpgbisgbk) · Medium · 0 upvotes · 0 comments - [AI voice agent for customer service: Where it works today](https://daily.dev/posts/ai-voice-agent-for-customer-service-where-it-works-today-rxx1zfsd5) · Netguru · 0 upvotes · 0 comments - [Customers prefer ChatGPT to company chatbots, Gartner finds](https://daily.dev/posts/customers-prefer-chatgpt-to-company-chatbots-gartner-finds-fzaszkirx) · The Next Web · 0 upvotes · 0 comments - [John Henry still leading the race vs. AI in customer service](https://daily.dev/posts/john-henry-still-leading-the-race-vs-ai-in-customer-service-jebmxi9pb) · The Register · 0 upvotes · 0 comments --- Tags: [#ux](https://daily.dev/tags/ux) [View this post on daily.dev](https://daily.dev/posts/ai--mj93nrhe9) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"你的AI客服學會的不是解決問題,而是讓用戶閉嘴","url":"https://daily.dev/posts/ai--mj93nrhe9","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/ai--mj93nrhe9"},"datePublished":"2026-07-14T07:28:13.957Z","dateModified":"2026-09-14T07:12:34.489Z","description":"Klarna's 2023 announcement of replacing 700 customer service agents with AI became a cautionary tale by 2025, when the CEO admitted quality had suffered and...","image":"https://media.daily.dev/image/upload/s--VDukGCjf--/f_auto/v1722860399/public/Placeholder%2002","thumbnailUrl":"https://media.daily.dev/image/upload/s--VDukGCjf--/f_auto/v1722860399/public/Placeholder%2002","isAccessibleForFree":true,"articleSection":"Medium","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Medium","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/medium","url":"https://daily.dev/sources/medium_js"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/ai--mj93nrhe9","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ux","timeRequired":"PT1M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Medium","item":"https://daily.dev/sources/medium_js"},{"@type":"ListItem","position":3,"name":"你的AI客服學會的不是解決問題,而是讓用戶閉嘴"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/ai--mj93nrhe9#faq","mainEntity":[{"@type":"Question","name":"What happened with Klarna's AI customer service after they replaced 700 human agents with AI in 2023?","acceptedAnswer":{"@type":"Answer","text":"By 2025, Klarna reversed course and began rehiring human customer service staff after CEO Sebastian Siemiatkowski admitted the company overemphasized efficiency and cost cutting at the expense of service quality, calling that approach unsustainable. The company adopted a hybrid model where AI handles simple queries while humans take over complex or emotionally sensitive cases. Anyone weighing AI versus human support tradeoffs can follow real-world case studies like this one on daily.dev."}},{"@type":"Question","name":"Why is deflection rate a misleading metric for measuring AI customer service success?","acceptedAnswer":{"@type":"Answer","text":"Deflection rate measures whether a conversation ended without being transferred to a human, not whether the customer's problem was actually solved. Analysis shows only around 14% of self-service interactions fully resolve issues, and even simple cases resolve just 36% of the time, yet companies commonly report deflection rates above 60%, meaning most of that gap reflects customers giving up rather than getting help. Teams designing support KPIs can dig into resolution-versus-deflection tradeoffs like these on daily.dev."}},{"@type":"Question","name":"What percentage of customers stop using a company after a bad AI customer service experience?","acceptedAnswer":{"@type":"Answer","text":"Close to 44% of customers who lose trust in AI customer service simply stop purchasing from that company altogether, often without complaining or explaining why, leaving retention numbers to quietly decline. Separately, a Qualtrics survey of over twenty thousand consumers found nearly one in five AI customer service users reported no benefit at all from the experience. Anyone tracking churn tied to AI support decisions can follow findings like these on daily.dev."}}]} ```