Stewart Golf's deployment of Crescendo represents a critical inflection point in how mid-market retailers approach AI-augmented support. The golf trolley manufacturer faced a structural problem endemic to growing B2C operations: a UK-based support team managing global demand across incompatible time zones, with response times stretching to 72 hours during peak periods. Rather than pursuing the conventional path of headcount expansion—which Hooper estimated would require a 50% increase—Stewart Golf evaluated native Zendesk AI capabilities and found them insufficient for the conversational quality customers expected. Crescendo's solution operated as a hybrid layer: AI handling 75% of routine enquiries within minutes whilst human agents managed escalations, with the platform providing weekend coverage that the internal team couldn't sustain. The results—one-minute response times and four consecutive weeks of 100% satisfaction—suggest the gap between platform-native AI and purpose-built conversational systems remains material, at least for brands prioritising experience consistency over cost reduction alone.
The implications cut across two distinct CX operating models. For teams already embedded in Zendesk or similar platforms, this case exposes a persistent weakness: knowledge base integration alone doesn't guarantee conversational competence. Stewart Golf's initial exploration of Zendesk's native capabilities failed because the platform optimises for ticket routing and resolution tracking rather than natural dialogue. Crescendo's ability to demonstrate a working assistant using existing knowledge base data within days suggests that third-party AI layers will continue capturing value from platform incumbents who treat AI as a feature rather than a foundational capability. The second implication concerns staffing strategy: if a 50% headcount increase was the alternative, then hybrid AI-human models become a retention tool as much as a cost lever, allowing existing teams to focus on complex problem-solving rather than repetitive triage. The question for larger platforms is whether this pattern—where customers outgrow native AI and seek specialist overlays—signals a structural market opportunity or a temporary gap that Salesforce's acquisition of Fin and similar consolidation moves will eventually close.
What remains unexamined is whether Stewart Golf's 100% satisfaction metric reflects genuine customer delight or the novelty effect of dramatically faster response times. The four-week window is notably short, and satisfaction scores often regress once customers normalise improved baselines. More substantively, the case demonstrates that conversational AI's real value lies not in cost displacement but in enabling service models previously impossible at scale—weekend coverage, sub-minute response times, consistent tone across channels. For CX leaders evaluating AI investments, the lesson is less about choosing between platforms and more about recognising that experience quality now requires deliberate architectural choices about where conversational intelligence sits in your stack.
Stewart Golf boosts customer support operation via AI powered customer experience platform Crescendo Retail Technology Innovation Hub