Salesforce集成CRM至Claude 用户无需打开原生应用即可操作

This article introduces Claudeforce, the new joint initiative between Salesforce and Anthropic, which has launched a Salesforce plugin embedded with Claude CoWork to deliver a new AI-integrated CRM user experience. Key takeaways are as follows:

1. Core features: The plugin comes with 37 pre-built sales capabilities covering common sales scenarios including meeting preparation, deal health reviews, and pipeline analysis. Users can query, update, and operate real-time CRM data directly without opening the native Salesforce application, drastically improving work efficiency. For example, a process that once required tens of thousands of clicks to filter pending opportunities can now be completed in just 30 seconds.

2. Usage and access rules: Enterprise administrators only need to complete a one-time connection to centrally manage authentication and access permissions. The plugin fully inherits users' existing permission rules, requiring no additional deployment, auditing, or per-account configuration, which greatly lowers the technical barrier for average users.

3. Rollout timeline: The product is currently available to a limited set of pilot customers. It will enter public testing in September, and will gradually expand to more business scenarios including service and marketing in the future.

This article offers multiple insights for brand owners on AI implementation, product R&D, and business model transformation. Key takeaways are as follows:

1. Industry trend insight: The enterprise software economy is gradually shifting from seat-based pricing to consumption-based pricing. As AI agents become the primary users of SaaS functionalities, per-user pricing models will gradually lose their rationale. Brand owners should monitor industry shifts early and prepare for business model transformation.

2. AI implementation insights: Most companies still struggle to convert AI pilots into measurable business gains, with core pain points being high deployment barriers and high management costs. Claudeforce addresses these frictions via pre-built capabilities, permission inheritance, and one-time deployment — a model well worth adopting for brand owners.

3. Partnership model insight: Deep integration between leading SaaS vendors and large AI model providers has emerged as a new trend. Following Salesforce's investment in Anthropic, the two companies have combined complementary strengths: this collaboration helps Anthropic reach millions of Salesforce sales users, while also boosting usage frequency for Salesforce. This cross-sector partnership model is a valuable reference for brand owners.

For sellers operating in the B2B sales space, this collaboration opens new opportunities for AI-driven efficiency gains while highlighting key risks to monitor. Key takeaways are as follows:

1. Efficiency improvement opportunities: The new plugin is built to assist sales teams with daily work, reducing workload from context switching and manual data aggregation. It can automatically generate daily work summaries, prioritize action items, flag key client changes, and build custom sales data dashboards, helping sales staff save significant time on repetitive tasks and focus on core conversion work.

2. Cost and pricing notes: The service is currently priced under a "headless consumption" model, where API call quotas are tied to users' existing Salesforce license tiers. Sellers need to purchase Claude inference services separately from Anthropic, as the two companies do not currently offer a bundled subscription. Sellers can evaluate costs based on their own usage to select a suitable plan.

3. Potential risk warning: Amid industry transformation, pricing power may gradually shift toward AI model providers, and existing seat-based pricing models will be gradually replaced. Sellers should monitor upcoming changes to pricing models and adjust their own cost structures early to adapt to industry shifts.

For factories looking to advance digital transformation and expand sales operations, this article offers valuable insights for digital implementation. Key takeaways are as follows:

1. Digital implementation insights: Factories commonly face high technical barriers, high multi-account permission management costs, and complex deployment processes when advancing AI-powered digital transformation, leaving many AI pilots unable to deliver actual business gains. Claudeforce's model of one-time deployment, permission inheritance, and pre-built functionalities is a valuable reference, effectively reducing implementation frictions and helping factories achieve AI-driven efficiency gains faster.

2. Tool application insights: As AI and enterprise services grow increasingly integrated, opening core data and workflows from SaaS vendors to AI models has become a new industry direction. Factories can leverage these new tools to integrate their own customer management and sales process data, improve internal sales team collaboration, and reduce internal management costs.

3. Cost and risk notes: Enterprise services are transitioning from seat-based pricing to API consumption-based pricing, allowing factories to purchase services flexibly based on actual usage needs and cut unnecessary seat costs. At the same time, factories should prepare in advance for the potential risk of rising costs driven by shifting pricing power to AI providers.

For service providers focused on enterprise services and AI implementation, this article covers the latest industry trends, core customer pain points, and referenceable solutions. Key takeaways are as follows:

1. Market opportunities and customer pain points: The enterprise generative AI market is growing rapidly: Gartner forecasts global end-user spending on generative AI models will reach $142 billion by 2025. However, most companies still struggle to convert AI pilots into measurable gains, with core pain points including high technical barriers to integrating AI into existing systems, high management costs for per-account permission configuration, and complex additional deployment and auditing processes. This represents a core entry opportunity for service providers.

2. Referenceable solutions: To address these pain points, Claudeforce leverages pre-built capabilities, permission inheritance, and one-time centralized enterprise deployment to effectively reduce AI implementation frictions. This solution can be reused across many enterprise AI implementation scenarios and is well worth adopting for service providers.

3. Industry development trends: Deep integration between SaaS vendors and large AI model providers is a new industry trend, and enterprise software pricing is shifting from seat-based to consumption-based models. Service providers can prepare for these shifts early and expand their businesses by leveraging the open ecosystems of leading vendors.

For enterprise SaaS platform operators, this article covers the latest industry developments, referenceable operational directions, and risks to avoid. Key takeaways are as follows:

1. Evolving user needs: Knowledge workers now widely want to reduce operational costs from switching between multiple platforms, preferring to complete data operations on original SaaS platforms directly within an AI interface, without switching back and forth between native apps. At the same time, enterprises want to reduce deployment and management costs for AI integration — these are core priorities for platform product optimization.

2. Referenceable latest operational practices: Salesforce launched Headless 360 to open its core data and workflows to AI models, turning organic user integration demand into an official product. It partnered with Anthropic to productize its mature internal solution, adopting a one-time connection model for centralized permission management that aligns with enterprise user needs, making it a valuable reference for platform operators.

3. Risks to avoid: During the transition to API consumption-based pricing, platforms may see existing seat revenue erode faster than API revenue grows. Additionally, as the interface layer gradually commoditizes, pricing power may shift to AI model providers. Platform operators should prepare for these challenges early and retain control over their own pricing.

For industry researchers, this article covers the latest industry developments in the integration of generative AI and enterprise SaaS, new business model experiments, and emerging research questions. Key takeaways are as follows:

1. Latest industry developments: Global leading SaaS provider Salesforce and leading large model developer Anthropic have expanded their partnership to launch the Claudeforce initiative, which embeds core CRM functionality directly into Claude. This is a benchmark project for deep integration between SaaS and large models, currently in pilot testing, with public launching coming in September, and gradual expansion to sales, service, marketing and other scenarios planned, making it highly valuable for research.

2. New business model exploration: This collaboration tests an entirely new pricing model, breaking away from traditional seat-based SaaS pricing to adopt a headless consumption model based on API calls, which aligns with the industry shift toward AI agents becoming the primary users of SaaS. The partnership also uses a split billing model with no bundled subscription, marking a new experiment in cross-sector collaboration.

3. New open research questions: This transition has also exposed new industry challenges, including the potential shift of pricing power to large model providers, the risk of rapid erosion of traditional seat revenue, and the persistent difficulty of AI implementation for most enterprises. This article includes Gartner forecast data, industry research findings, and internal Salesforce practice data that can serve as foundational material for further research.