Razorfish Offers Clients Real-World AI Use Cases Through Beta Labs

Publicis agency Razorfish has widely opened up its suite of gen AI tools Beta Lab, to give all brand clients practical examples of how can be used to enhance content creation and improve customer experiences. Beta Lab has been open to over half of the agency's clients—including Samsung—since this summer. The lab has already featured in many Razorfish pitches and has helped secure deals, the agency said, but wouldn't say how many. “Brands that are testing this are seeing the biggest impact against scale—the ability to do more and to supercharge insight development and content creation,” said Cristina Lawrence, evp of consumer and content experience, Razorfish. “At a macro level, this is allowing us to plan and produce creative campaigns which wouldn’t have been possible with more traditional creative production processes.” The lab is part of Publicis Groupe’s commitment of investing north of $300 million on AI, and is compatible with Publicis’s main AI platform Core AI. 61% of agencies are currently using AI, according to Forrester, spurring other holding companies, like Havas and its $429 million investment, to ramp up their AI commitments. Agencies like Monks and Supernatural are also exploring innovative applications such as AI-powered focus groups. At the heart of Beta Lab is the Conversational Intelligence Platform (CIP), which helps brands develop custom AI chatbots. Beta Lab also includes the Virtual AI Content Studio—a creative platform launched last year equipped with digital production, visual effects, and AI tools. This studio lets clients streamline image and video creation for ads, generating on average 10 times more assets, Razorfish claims, compared with not using AI. Beta Lab is built on large language models like OpenAI’s GPT-4o and Google’s Vertex, with integrated APIs and partnerships with data vendors. The agency spent the last year and a half developing it in response to growing client interest in gen AI. The services are billed as part of client expenses. “Each iteration [of gen AI models] dispels the hype and demonstrates not only this tech’s potential profound impact on everything but also where we can see real, tangible benefits against the use cases of today," Lawrence said. How brands are using the lab An automotive client has used CIP to add a concierge-like chatbot to its site. The chatbot tailors the browsing experience based on what users search for, such as vehicle specs or model comparisons. “It’s like search in that you ask a question, but it pulls and serves relevant information like copy and imagery,” said Lawrence. The chatbot aligns with the brand’s tone of voice, ensuring responsive interactions that reflect the overall messaging. Other clients are also exploring Beta Lab’s custom persona AI chatbots, developed through a feature called SEED (Synthetic Environment for Engagement and Dialogue). These personas use the brand’s first-party data and third-party sources like media partners and segmentation data. For instance, an online payment brand developing a mobile travel site is testing these personas due to the inability to ask focus groups targeted questions about desired features. “The use case here is not to replace traditional market research but to supplement it and extend it to get to granular insights that helped the team force rank the feature sets,” said Andrew McKernan, gvp, content & consumer experience strategy, Razorfish. Samsung has used the Virtual AI Content Studio to generate relevant AI-driven ads at scale for its SmartThings ecosystem, opting for timely and budget-conscious production. Instead of a traditional photoshoot, which would have taken significantly longer, Razorfish produced the assets within a few days. “A traditional shoot of this magnitude would take five to 10 times longer, to execute,” said Lawrence. While not a complete substitute for conventional shoots, Lawrence emphasized that AI tools open “new doorways for creativity,” especially when time and budget constraints are a factor. Many clients at Razorfish are still navigating their internal policies on gen AI, with about a 50/50 split between those eager to embrace it and those more hesitant. According to Lawrence, clients in non-regulated industries are more likely to adopt AI, while those in regulated sectors remain cautious due to governance concerns. “This technology has outpaced society’s ability to regulate it,” she said.

AI News

Razorfish Offers Clients Real-World AI Use Cases Through Beta Labs

The lab has already featured in pitches and won new business

Trishla Ostwal

About

Subscriptions

Events

Publications

WORK SMARTER - LEARN, GROW AND BE INSPIRED.

Subscribe today!

Trishla Ostwal