CRM Software Faces Disruption From AI Agents

Customer Relationship Management (CRM) software is facing a fundamental disruption driven by the emergence of AI agents capable of consolidating disparate data sources and fostering a unified understanding of individual customers. This shift challenges the traditional CRM model, which is built on fragmented data residing in product logs, billing systems, support queues, and manual representative notes. These systems create records but fail to deliver true customer insight, a problem exacerbated by customer base growth. During a four-year period at HubSpot, the customer base expanded from approximately 70,000 to over 220,000, a scale that rigorously tests existing assumptions about customer data management. The current CRM architecture, characterized by disjointed data that is difficult to integrate meaningfully, is unlikely to survive in its present form.

The evolution of the customer journey from a funnel to a flywheel and then a loop has highlighted a persistent defect: these models focus on the company's internal processes and sequence of actions, rather than the actual customer experience. A customer three months into a contract might encounter issues like failed logins and unresolved billing disputes, while simultaneously receiving an upgrade offer from a representative unaware of these problems. This fragmentation is reflected in customer sentiment, with Salesforce's State of the Connected Customer survey indicating that 55% of customers feel they are interacting with separate departments rather than a cohesive company, and 56% report having to repeat information to different representatives. Historically, this disconnect was manageable because software primarily served as an information delivery tool to human agents, who could only process information from a single screen at a time.

However, the advent of agentic AI is fundamentally altering this dynamic. Tech companies are now focused on leveraging AI agents to transform siloed data into structured databases and developing tools that can efficiently locate and process this information to generate relevant answers. This enables a paradigm shift where clients are billed for the outcomes and resolutions achieved, rather than for access to the software itself. For instance, in customer support scenarios, clients may pay per issue resolved instead of per user seat. This new model, termed 'Service as Software' by Foundation Capital's Jaya Gupta and Ashu Garg, represents a potential $4.6 trillion market opportunity. The customer support sector has been an early adopter of Service as Software due to the inherent nature of resolving customer issues, but its application is expected to broaden across various business functions as AI capabilities mature.