How Uber ensures every agent meets the quality bar before they ever touch a customer

As Uber expanded into new markets, their sales growth team faced a quality control problem. New agents were going live with customers before they had fully mastered the product pitch, objection handling, and compliance requirements specific to each region. The existing QA process relied on manager ride-alongs and spot checks — methods that didn't scale and left gaps in coverage. The team needed a way to objectively verify that every agent was ready before their first customer interaction.

Uber implemented Jam's AI Sales Simulations as a mandatory certification step before agents could go live. Each agent completes a series of realistic scenarios tailored to their specific market — covering everything from initial outreach to complex objection handling. Agents must hit defined competency thresholds across multiple dimensions before they're cleared. Those who fall short receive targeted practice on their weak areas and can re-certify when ready. Managers get a clear dashboard view of team readiness across all regions.

The impact on quality was immediate. With every agent now certified before launch, early-stage customer escalations dropped by 50%. Agents reached full readiness 3x faster than with the previous shadowing and ride-along approach. The certification data also gave leadership visibility into which markets and skill areas needed the most support — enabling proactive investment in training before problems surfaced in customer interactions.

"We now know agents meet our quality bar before they ever touch a customer. That's the difference." — Gijs Meeder, Sales Growth Expansion