The « Vestiaire Collective’s » smartbot
 

The context

To scale customer support across Europe (EN, FR, IT), we needed to transition from manual ticket handling to an intelligent automated solution. The global inbound contact rate was high, and the objective was to reduce live ticket volume without compromising the customer experience.

The purpose

The project aimed to transform customer support through three pillars:

  • Intent Detection: leverage GenAI (uGPT) to automatically categorize issues with an 80% accuracy goal

  • Automated Resolution: focus first on low-volume/high-complexity issues (Fraud, Account) before expanding to general queries to reach a 5% global resolution rate.

  • CX Excellence: maintain high satisfaction by monitoring CSAT and NPS comments to iterate on bot flows.

The results

The rollout delivered immediate operational impact:

  • Scale: successfully launched in EN and FR, covering ~70% of total messaging volume.
     

  • Efficiency: automated 22 contact reasons within the first month, handling 8.8k weekly conversations.
     

  • Performance: achieved a 32% deflection rate in automated workflows and contributed to a -8.1pp decrease in the global inbound contact rate.
     

  • Satisfaction: maintained a 85% CSAT through continuous flow optimization

 
 
Retour en haut