Five9 surveyed consumers and business leaders in the US, UK and Germany. Only 46% rated their last AI chatbot experience positively. Two-thirds rated live phone, email or in-person support positively. Nearly half who switched to a human did so because the bot misunderstood them. Customers often speak in ways bots cannot parse. Context and knowledge bases can be outdated. Organizations should test bots with real, messy conversations. They should also make human escalation easy. Five9 found trust in AI rises when a human option exists. CX leaders should track failure and escalation rates.

• 46% positive AI chatbot experiences
• 66% positive live phone/email/in-person experiences
• 46% abandoned AI due to misunderstandings
• 3,000 consumers surveyed