Tower, a New Zealand insurance provider, deployed Amazon Connect's AI capabilities in a pilot programme and achieved measurable reductions in call handling times. The initiative demonstrates that contact centre AI isn't merely a cost-reduction theatre—when properly integrated into existing infrastructure, it delivers tangible efficiency gains that directly impact agent productivity and customer throughput. This matters because Tower's results validate what many CX leaders have suspected: the bottleneck isn't whether AI can help, but whether organisations can implement it without disrupting established workflows or alienating agents who fear displacement.
The strategic question emerging from Tower's pilot is whether mid-market organisations should prioritise Amazon Connect specifically, or whether the real competitive advantage lies in any vendor's ability to integrate AI seamlessly into their existing stack. Tower's choice of Connect suggests AWS's contact centre solution is maturing beyond its early-adopter phase, but the broader implication is that CX teams now face a different decision matrix than they did two years ago. Rather than debating whether to adopt AI at all, teams must evaluate which platform—whether Connect, Salesforce's Agentforce, or native capabilities within Zendesk and Freshdesk—offers the fastest path to measurable call time reduction without requiring wholesale platform migration. For organisations already entrenched in competing ecosystems, the question becomes whether switching costs outweigh the efficiency gains Tower has demonstrated.
What's notably absent from Tower's narrative is any discussion of call quality, customer satisfaction, or whether shorter calls actually improved outcomes. Call time reduction is a lagging indicator that's easy to measure but potentially misleading; if AI is simply rushing agents through interactions, the long-term cost of rework and churn could erase the efficiency gains. CX leaders evaluating similar pilots should demand metrics beyond speed—first-contact resolution rates, customer effort scores, and repeat contact rates—before committing budget to platform shifts or expanded AI deployments.
Tower cuts call times with Amazon Connect AI pilot IT Brief New Zealand