你现在的竞争对手，是每天工作 12 小时、全年无休的 AI 创业公司，且部分业务已经由 AI 提效。 在 AI 时代，企业就没有「不转型」的选择。 这是在回顾 Intercom 从 SaaS 传统老牌公司到 AI-first 的转型过程中，其创始人 Eoghan McCabe 给出的 答案。 CRM 代表企业 Intercom 的 AI 转型故事堪称传奇。Intercom 的传统业务曾估值数十亿美元，年度经常性 收入（ARR）达数亿美元，但经历了净新增 ARR 连续五个季度下滑，净新增收入几乎为零的"灾难"时 期。 内部缺乏决策力，战略方向也显得漂移不定，整个组织陷入了一种"舒适"但危险的惯性之中 。 转型成为急迫但又棘手的事。在 Eoghan McCabe 重新接手后，迅速做出了一系列调整：裁员、砍掉其 余业务，聚焦客服领域，快速开发出客服 AI Agent 产品 Fin......曾经备受诟病的定价问题（过高且不透 明的定价），也大刀阔斧地调整为「按结果付费」，用 99 美分解决一个问题。Eoghan 认为，定价应该 基于「价值」而非「成本」 ，而成本是企业自己要解决的问题。 如今，Fin 已经逐步取代传 ...

在瞬息万变的科技行业，每天都有企业在创新的赛道上奋力奔跑，也有不少公司因跟不上技术迭代的步伐而陷入生存困境。有的企业在市场的冲击下逐渐 沉寂，有的则在绝境中苦苦寻觅破局之路。 对讲机领域曾一度因技术瓶颈和市场需求变化，让不少从业者感到迷茫，许多公司面临着转型无门、业绩下滑的严峻挑战。在这样的大环境下，有一家对 讲机企业却上演了一场令人惊叹的逆袭大戏 —— 创始人决定孤注一掷押宝 AI，让濒临破产的公司创下业绩高峰，如今市值破百亿，超过大多数软件公 司。 Intercom 是一个客户服务平台，成立于 2011 年，总部位于美国旧金山，Eoghan McCabe 是公司联合创始人兼 CEO。 Eoghan McCabe 在爱尔兰出生长大，1996 年他在美国在线 (AOL) 上建立了自己的第一个网站，并于 2000 年创办了自己的第一家互联网公司，为只有 1 万 人口的家乡打造了一个互联网门户网站。在都柏林圣三一学院学习计算机科学期间，他进一步拓展了自己的抱负。在此期间，他仔细研读了 37signals 出 版的关于软件的关键著作，梦想着像他们的 Basecamp 一样创办自己的公司。 2006 年，Eoghan ...

作者 | 冬梅 在瞬息万变的科技行业，每天都有企业在创新的赛道上奋力奔跑，也有不少公司因跟不上技术迭代的步伐而陷入生存困境。有的企业在市场的冲击下逐 渐沉寂，有的则在绝境中苦苦寻觅破局之路。 对讲机领域曾一度因技术瓶颈和市场需求变化，让不少从业者感到迷茫，许多公司面临着转型无门、业绩下滑的严峻挑战。在这样的大环境下，有一家 对讲机企业却上演了一场令人惊叹的逆袭大戏 —— 创始人决定孤注一掷押宝 AI，让濒临破产的公司创下业绩高峰，如今市值破百亿，超过大多数软件 公司。 Eoghan McCabe 在爱尔兰出生长大，1996 年他在美国在线 (AOL) 上建立了自己的第一个网站，并于 2000 年创办了自己的第一家互联网公司，为只有 1 万人口的家乡打造了一个互联网门户网站。在都柏林圣三一学院学习计算机科学期间，他进一步拓展了自己的抱负。在此期间，他仔细研读了 37signals 出版的关于软件的关键著作，梦想着像他们的 Basecamp 一样创办自己的公司。 2006 年，Eoghan 大学刚毕业，创办了自己的软件开发公司，并命名为 Eoghan McCabe Ltd.。该公司的第一款软件产品是一款名为 Fo ...

Core Viewpoint - AI is fundamentally changing the pricing logic of software, shifting from traditional seat-based pricing to usage-based or outcome-based pricing models [2][11][20]. Group 1: AI Pricing Transformation - The traditional seat pricing model is becoming less viable as AI increases efficiency, leading to fewer users and a need for new pricing strategies [11][12]. - Implementing usage-based pricing faces challenges such as the need for real-time billing systems, dynamic pricing models, and maintaining large volumes of accurate data [3][15]. - Pricing models for AI products can be analyzed based on attribution capability and autonomy, with stronger attribution and autonomy leading to greater pricing power [32][36]. Group 2: CEO Considerations for Pricing Transition - CEOs must focus on sales compensation structures and the division of sales responsibilities when transitioning to usage-based pricing [3][22]. - A hybrid business model, combining seat pricing and usage-based pricing, is expected to dominate in the coming years, especially for application-level products [3][13]. - The sales team's role must evolve to ensure that actual usage aligns with revenue recognition, avoiding the pitfalls of recording false revenue [22][23]. Group 3: Challenges in Implementing Usage-Based Pricing - Real-time monitoring is essential to manage the risk of unlimited spending in usage-based pricing models, as seen in cases like Segment [15][16]. - The dynamic nature of pricing models complicates the creation of a universal billing engine, as contracts often vary significantly [15][16]. - Maintaining a reliable data chain is crucial for accurate historical data storage, which is necessary for future pricing adjustments [15][16]. Group 4: Strategic Importance of Usage-Based Pricing - Usage-based pricing directly ties revenue to the value created for customers, allowing for a more flexible and responsive business model [17][20]. - Sales commissions in usage-based models must be adjusted to align with actual product usage, preventing cash flow mismatches [18][22]. - The integration of value creation across departments is essential for the success of usage-based pricing, requiring a shift in company culture and operations [19][21]. Group 5: Future of Pricing Models - The trend is moving towards a mixed pricing strategy, with a significant portion of companies expected to adopt outcome-based pricing in the next few years [37][49]. - Companies must enhance their products' autonomy and attribution capabilities to unlock greater commercial value [37]. - The evolution of pricing models reflects a broader shift in the industry, where agility and adaptability are key to maintaining competitive advantage [43][49].

Core Viewpoint - Venture capitalists are increasingly investing in startups that develop AI for customer support automation, with companies like Intercom and Kore.ai showing strong revenue growth potential and facing competition from established players [1][4]. Group 1: Intercom - Intercom is negotiating to sell over $100 million in shares, potentially achieving a valuation of over $2 billion [1]. - The company’s last equity financing was in 2018, with a valuation of $1.15 billion, and it has recently launched an AI customer support assistant, Fin, which is expected to contribute significantly to revenue growth [2]. - Intercom's annual recurring revenue was approximately $300 million at the end of last year, with Fin's contribution still small but rapidly growing [2][4]. - The company has seen its net revenue retention rate increase from 112% to 146% since adopting a pay-for-performance pricing model [4][5]. Group 2: Kore.ai - Kore.ai is in talks to raise about $150 million at a valuation between $2.5 billion and $4 billion, with an annual recurring revenue of approximately $110 million [1][3]. - The company expects its annual recurring revenue to grow by up to 80% to around $200 million by next March, with a gross margin of 80% on its subscription business [3]. - Kore.ai's consulting business accounts for 10% of its total revenue, and it primarily serves large clients in the banking and insurance sectors [5]. Group 3: Competitive Landscape - Both Intercom and Kore.ai face intense competition from newer AI startups like Sierra and Decagon, as well as established customer service software providers such as Zendesk and Salesforce [4]. - The AI customer support sector is seeing significant investment and innovation, with a focus on technologies that can autonomously resolve customer issues and enhance human agent efficiency [4].