L’Oréal Cuts Customer Service Time 64% With AI
A L’Oréal adviser who needed to issue a customer a discount code once clicked through about 15 screens over 10 minutes. Now the adviser types a prompt, and artificial intelligence generates the code in about 1 minute, Salesforce reported. Across 25 North American brands, that kind of help has cut average case handling time 64%, […] The post L’Oréal Cuts Customer Service Time 64% With AI appeared…
L’Oréal has implemented artificial intelligence to cut average customer service case handling time by 64%, according to a Salesforce case study. Previously, a L’Oréal adviser would issue a discount code by clicking through 15 screens over 10 minutes. Now, an AI-powered agent, called Agentforce, can generate the code in just one minute after receiving a prompt from the adviser.
Across 25 North American brands, the AI tool has resulted in a 64% reduction in average case handling time. In addition to L’Oréal, YSL Beauty and Kiehl’s have incorporated AI agents that can close order status requests independently on their websites. L’Oréal aims to resolve 60% of inquiries without human intervention. AI agents summarize each conversation as it arrives and draft replies in a brand’s voice, such as Lancôme or CeraVe.
Advisers review and send the drafted replies, with the AI checking that the logged reason matches the conversation before closing. More than 300 advisers use the tool daily, with 100% adoption among eligible staff. L’Oréal plans to remove the need for advisers altogether by having the AI handle more complex consumer requests. The AI agents are limited to preapproved language and cannot write new sentences.
If a request falls outside the agent's scope, the chat is transferred to an adviser with the full history attached. L’Oréal has limited the AI agents to handle money-related tasks, with only 24% of Americans comfortable having an agent shop and pay for them.
Written by urgent.news from PYMNTS's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.