Zoom Expands AI Companion Features for Conversational Intelligence in Sales

Zoom pushed conversational intelligence deeper into everyday B2B sales calls this week, rolling out advanced coaching features to its AI Companion. But for UK SMEs, buying these tools without fixing the underlying sales process will only help reps fail faster and with better transcripts. The drive to automate feedback across Zoom's massive user base highlights a growing gap between buying AI and actually closing deals.

Zoom expands AI Companion sales features

Zoom is heavily pushing its conversational intelligence capabilities, integrating automated meeting summaries and feedback directly into its AI Companion. The tool now processes sales calls to extract next steps, highlight talk tracks, and provide immediate post-meeting coaching notes to reps.

Instead of paying for expensive third-party add-ons, companies now have conversational intelligence built directly into their default video platform. Zoom claims the companion helps users save time and improve meeting outcomes by analysing talk-to-listen ratios and question rates. This matters because Gartner predicts the vast majority of B2B interactions are now entirely digital, making video calls the primary battleground for closing deals.

The feature set is impressive, but it creates a new problem. Sales directors are suddenly flooded with data about how their teams speak, what objections they face, and where calls fall apart. The software highlights the exact moment a prospect loses interest, but it does not tell a struggling rep how to fix a fundamentally flawed pitch.

Why bad scripts break good AI

The trap for a 50-person UK business is thinking that conversational intelligence is a substitute for a strong sales process. It isn't. If your core value proposition is weak, AI will just give you highly accurate data showing exactly how badly it lands.

I see too many sales directors turn on these features expecting an immediate revenue bump. Instead, they get a dashboard of metrics... talk time, filler words, and competitor mentions... that distracts from the real issue. If your reps don't know how to qualify a lead or handle a standard pricing objection, telling them they spoke for 62% of the call is useless.

AI call coaching amplifies whatever you feed it. Feed it a tight, tested sales script, and it will help you refine the margins. Feed it a chaotic, unstructured pitch, and it will just document the chaos. The real risk here is distraction. Teams spend hours reviewing AI-generated coaching notes instead of doing the hard work of rewriting their discovery questions or tightening their objection handling.

Three steps to take before turning on AI coaching

1. Audit your current script. Before you look at a single AI transcript, write down your baseline discovery questions and objection responses. If you don't have a documented baseline, AI can't tell you if reps are deviating from it.
2. Pick one metric to fix. Don't look at the entire AI dashboard. Choose one specific behaviour to track this month, like how often reps ask for the budget on the first call. Ignore the rest of the noise.
3. Coach the process, not the transcript. Use the AI summaries to spot structural flaws in your pitch, not just to correct individual reps. If three different reps hit a wall at the same point in the call, the script is broken, not the salesperson.