Cars24 recovers 12% of lost leads with OpenAI voice agents at scale

The 12% lead recovery figure is the number worth scrutinizing: it implies Cars24 had a measurable, pre-existing leakage problem that AI addressed, which means the ROI case here is unusually concrete compared to most enterprise AI deployments that report efficiency gains without a clear counterfactual baseline.

Cars24 is essentially a live case study for the argument OpenAI's European go-to-market team has been making in parallel. The 'From AI Pilots to Production' coverage from the same day identifies three conditions for scaled deployment: measurable ROI, executive sponsorship, and treating adoption as business transformation rather than a technology project. Cars24 checks all three, and the multi-team internal workflow integration maps directly onto what Emmanuel Marill described in 'Think AI-First' as reconceiving operations around AI rather than retrofitting models onto legacy processes. What's notable is that Cars24 did this in a high-volume, low-margin transactional context, which is a harder environment than the knowledge-work use cases most enterprise AI case studies feature.

Watch whether Cars24 or comparable automotive marketplaces in emerging markets publish churn or conversion data tied to these agents within the next two quarters. If lead recovery rates hold or improve at scale, it validates the agentic infrastructure model for transactional verticals broadly; if they plateau, the 12% figure may reflect one-time recapture rather than a durable operational advantage.