AI Calling Systems That Actually Connect: Fixing Answer Rates with Kevin DeMeritt

This episode of Dialed In digs into why so many voice AI calling programs fail before a conversation even starts. Host Rachel sits down with Kevin DeMeritt of 2X to unpack the gap between flashy AI agents and the less glamorous reality of connection rates, caller reputation, and compliance. Listeners will learn why focusing on how human an AI sounds is not enough, how deliverability and trust drive real outcomes, and what operational systems need to be in place for AI-assisted outbound calling to actually work at scale. The discussion covers practical tactics, from pre-call text sequences to local presence strategies, and explains how they translate into measurable answer-rate improvements. - Voice AI misconception: Why teams obsess over conversation quality while ignoring whether anyone is actually answering their calls. - Connection rate pitfalls: How spam-likely flags, poor number reputation, and bad timing quietly destroy outbound performance. - Caller credibility metrics: What Kevin looks for in audits and how psychology, number reputation, and sequencing drive connectability. - Pre-call text and local presence: How short, human-style texts and local numbers can lift answer rates by 10–20%. - Operational complexity: Why DIY AI calling often fails without automation for texts, voicemails, compliance, and smart human-transfer timing. The conversation closes with a look at the future of AI calling: a shift toward connectability over voice polish, preserving momentum with 45–120 second AI interactions, and preparing for tighter regulatory oversight as AI-driven conversations become more common. If this episode surfaced gaps in your own outbound or compliance strategy, subscribe and share it with your legal or operations team. To see how compliance-first AI calling works in practice, visit 2x.ai or email your questions to podcast@2x.ai.