How E.l.f. Beauty is building AI into its community management

E.l.f. Beauty's AI tool E.l.f.luencer, trained on the company's stylebook and years of human-written social comments, has raised its social media comment response rate from 30%-40% to above 80%, with 90% of comments generated by AI in the brand's tone, according to Chief Technology and AI Officer Ekta Chopra. The company is piloting 85 agentic AI use cases and has created internal tools including the chatbot B.F.e.l.f, used by more than 80% of employees, and an IT help desk assistant called e.l.f.line.

E.l.f. Beauty’s posts on TikTok https://www.fastcompany.com/section/tiktok and Instagram receive hundreds if not thousands of comments when they go online. Often, the comments include questions about when a product is launching, how the product compares to other E.l.f. offerings, or what type of skin or hair it’s appropriate for. To answer shoppers’ questions in real time, the company created an AI https://www.fastcompany.com/section/artificial-intelligence tool called E.l.f.luencer—trained on the company’s stylebook and years of human-written social comments—that autogenerates responses to comments left by consumers on social media platforms. The comments are approved by human community managers before being posted. “The tool saved a lot of time for community managers ,” says Ekta Chopra, the company’s chief technology and AI officer. “Our response rate was like 30%-40% before launching E.l.f.luencer. Fast forward, it’s gone above 80%; 90% of the comments are generated by AI in our brand tone, using the language and emojis E.l.f is known for. ” Chopra adds that artificial intelligence at the company is used to augment productivity https://www.fastcompany.com/section/productivity , but all outputs are closely reviewed by human employees. This is just one initiative Chopra, a 10-year E.l.f. veteran, has overseen. In her role, Chopra helps the company identify and implement its AI ambitions. To oversee AI implementation across the company, Chopra created a cross-functional AI steering committee spanning legal, marketing https://www.fastcompany.com/section/marketing , and R&D with the goal of moving most agentic AI pilots into full production within six months and minimizing risks associated with the technology. Under her leadership, the company is piloting 85 agentic AI use cases. Her deep investment in using AI across departments has helped E.l.f create internal tools and consumer-facing tools like E.l.f.luencer. Internally, Chopra has created AI tools to streamline processes and distribute information at every level of the company. There’s an internal chatbot called “B.F.e.l.f” that is used by more than 80% of employees. There’s also an IT help desk assistant called “e.l.f.line” that the company says can accomplish support work equivalent to that of one human help desk employee. Additional internal applications include tools to help products get surfaced by large language models when people search anything on the internet, as SEO continues to deliver fewer returns. To accomplish all of this, Chopra created a secure internal AI platform available to every employee that provides approved connectors and APIs, allowing nontechnical staff to build their own agents or apps and deploy them without relying on a central tech team. “We want to democratize that access for elves E.l.f. employees because they’re closest to the work and they can really reimagine that work,” Chopra says. Among the tools employees have built is a dashboard that tracks social sentiment across campaigns and products.