Text's deployment of AI selling agents within its live chat platform represents a fundamental repositioning of customer service from cost centre to revenue driver. The early results from their 600-company pilot are striking: a 266% increase in conversion rates, 39% uplift in chat sales attribution, and 60% growth in agent engagement within two months. These figures suggest that the traditional separation between support and sales functions is collapsing. Rather than routing qualified leads to sales teams, AI agents now identify purchase intent in real time, recommend products, surface promotions, and complete transactions autonomously—operating around the clock without additional headcount. The mechanism is straightforward: agents ingest behavioural signals, purchase history, and in-session context to recognise moments when intervention increases conversion probability, then act on that insight at scale.
The implications for CX teams are substantial but require careful navigation. Support leaders must confront a strategic question: if customer service conversations now generate direct revenue, how should team structures, KPIs, and training programmes evolve? The traditional resolution-speed metric becomes insufficient when agents are simultaneously closing sales and handling escalations. Text's response—introducing an AI Supervisor role that requires analytical thinking, customer psychology expertise, and the ability to translate business strategy into agent behaviour—signals that the skill set required of support professionals is shifting upward. This creates both opportunity and risk: teams that upskill to manage and refine AI agents gain influence over revenue outcomes, whilst those that treat agents as simple automation tools risk becoming obsolete. For administrators running platforms like Zendesk or Freshdesk, the question becomes whether your current infrastructure can support this dual mandate, or whether purpose-built solutions like Text will fragment the support stack further.
The human element remains non-negotiable, though its role has transformed. Text's emphasis on seamless handoffs—where agents recognise when a conversation requires human judgment and escalate before the customer requests it—reflects a maturity that many early AI implementations lack. The free AI Supervisor training programme signals that Text understands this transition requires investment in people, not just technology. For support teams already managing agent-assisted workflows, the challenge is distinguishing between genuine escalation criteria and moments where human involvement actually reduces conversion. The data suggests that well-trained AI agents operating within clear business rules outperform human-only models on resolution rates (averaging 74%, reaching 90% in some cases), yet the 39% sales attribution increase implies that human agents still close higher-value deals or handle complex negotiations. The competitive pressure is real: vendors offering similar capabilities will proliferate, and teams that fail to operationalise AI supervision will find themselves managing increasingly autonomous systems they no longer fully understand.
Customer Service As a Profit Engine: Text’s New AI Selling Agents Drive 266% Conversion Rate Increase and 39% Live Chat Sales Attribution The AI Journal