L'Oréal handles consumer support cases 64% faster with Agentforce

AI inside every case

To help advisors spend less time on routine tasks, L'Oréal brought agentic AI into Service Cloud, the workspace advisors already use to manage consumer cases.

As each case arrives, Agentforce goes to work alongside the advisor. It summarizes the conversation so far, so the advisor picks up the thread without reading back through it, and confirms what the consumer is actually asking for before work begins.

For the advisor, routine tasks now happen inside the same workflow. Generating a promotional code, for example, is now as simple as entering a prompt. Agentforce handles the whole process in about 1 minute.

Agentforce also drafts the first version of consumer emails using the appropriate voice for each L'Oréal brand. Advisors review the draft, make any needed edits, and send the response. And before a case is closed, Agentforce validates that the selected case intent matches the conversation, helping improve consistency across the organization.

Automating these repetitive tasks helps advisors move through cases more quickly while keeping people in control of conversations that require judgment, empathy, or exceptions.

“We don't want to replace our reps. We want to augment them,” said Alberto Rodrigues, AVP of Consumer Care Innovation & Intelligence for L'Oréal USA.

Grounded in business data

Agentforce is embedded within the Agentforce Service Console, allowing advisors to work with AI in the same interface they already use to manage consumer cases. They don't have to switch between applications or copy and paste information into a separate tool.

To give advisors the full context behind each interaction, L'Oréal grounds Agentforce in their own business data through Data 360, which brings together case history, CRM records, and knowledge articles from Service Cloud alongside order information from SAP S/4HANA.

That context does two jobs at once. Approved knowledge tells Agentforce what's accurate to say. Consumer records tell it who's asking and what they've already bought or asked about. Working from current data means advisors aren't reconciling a stale snapshot against what a consumer is telling them — a promotion that changed this morning, an order that shipped an hour ago.

The Atlas Reasoning Engine then determines the next best action for each case — summarizing what's happened, drafting a reply, or flagging when the selected case intent doesn't match the conversation. Brand-specific prompt instructions keep every response in the right voice, and each conversation draws on that brand's own knowledge and case history.

Proven pilot, confident expansion

Agentforce has reduced average case handling time by 64%, cutting the time to resolve a case from about 7 minutes to about 2.5 minutes. Individual tasks improved even more — generating a promotional code, which once meant navigating about 15 screens over roughly 10 minutes, now takes about 1 minute. Today, more than 300 advisors use Agentforce every day, and adoption has reached 100% among eligible advisors.

L'Oréal’s deliberate deployment approach enabled them to learn as they went. “We started with a pilot in Canada with a single brand, which allowed us to learn which intents we could operate on,” said Luc Antoine, Global Director of Innovation, Consumer Experience. “This set us up for success across the 25 brands deployed today in North America.” Europe is the next region to come online. 

Starting with employee-facing use cases also helped L'Oréal build a strong foundation for future agentic experiences. Advisors who worked with Agentforce became advocates for the technology, and business leaders have already identified new opportunities, including agentic order support and hardware troubleshooting.

“You can set out to build a unicorn, but we weren't going for perfection — otherwise this would've been a launch two years from now,” said Antoine. “We take a crawl, walk, run approach very seriously and it’s serving us well.”