Lacoste put a conversational AI agent, Ask René, into production on lacoste.com US four months after the project started: two months of design and two months of development and testing. Four days after launch on the US market, Converteo reported consumer satisfaction above 4/5, an escalation rate on track for -30% compared to the previous chatbot, and handling time for complex cases under 8 minutes on the human advisor side.
The previous chatbot on lacoste.com was purely deterministic and limited to after-sales service, unable to cover product discovery, recommendation or purchase assistance. Lacoste's teams wanted a digital advisor able to reproduce, at scale and in natural language, what the best in-store salesperson does, while respecting the brand's codes, down to the wording used when escalating to a human advisor.
The project kicked off in November, with the constraint that the existing chatbot on the US market had to be replaced by February-March. Converteo, in partnership with Google, delivered the platform, and its teams worked on Lacoste's premises within the brand's digital factory as a shared agile squad.
The solution uses Google's Gemini Enterprise for CX, which comes with pre-configured agents and customer experience building blocks. The agent is connected to Salesforce Commerce Cloud for the catalog and e-commerce layer, to Salesforce Service Cloud for ticket management and escalation to advisors, and to Manhattan Associates for order tracking and logistics. A central orchestrator agent routes each request to the most appropriate expert agent based on detected intent and the criticality level of the request. The knowledge agent answers frequently asked questions, such as general terms and conditions, return policy and pre-sales questions, from an enriched and trained FAQ base. The order tracking agent handles parcel tracking, cancellation requests and delay management, interfaced with Manhattan Associates, Lacoste's OMS tool, to retrieve real-time information. The product quality agent manages returns related to defects or post-wash issues, qualifying the issue and preparing the case for a human advisor. When escalation is required, the advisor retrieves the entire conversation, including any photos shared by the customer, directly in Salesforce Service Cloud with no additional data entry.
Training the language model on Lacoste's brand-voice constraints (empathy, short sentences, avoiding overly verbose phrasing, rapid escalation in case of tension) required careful prompting and fine-tuning work, which is ongoing. Deployment continues with Canada next, followed by Europe and Latin America, as teams adapt the agent to each market's carriers, local offers and languages. Next identified workstreams are personalized product recommendation, conversational personal shopping, multimodal interaction (voice, image) and omnichannel integration with in-store salespeople.