How beauty brands plan to embrace consumer-facing AI in 2024

“If you’re a marketeer, you must be an expert in gaming, technology, the Metaverse, you name it — every year. If you want to be a disruptive brand, you have to be knowledgeable and an early adopter and lean in,” said Nora Zukauskaite, global marketing director for Brand Agency London, which houses makeup brands Ciaté London and Lottie London and Gen-Z beauty brands Skin Proud, Hair Proud and Body Proud.

Glossy spoke with Zukauskaite, along with other brand executives and AI leaders from consumer-facing companies to understand the current use cases of AI for the beauty industry and the consumer-facing applications that are coming in 2024.

Brand Agency London:

Starting in early May, Brand Agency London began to employ generative artificial intelligence in marketing, communications and product development using ChatGPT. ChatGPT has aided press release writing, social media caption writing and competitive brand research.

Zukauskaite said there is a caveat: Employees still need to cross-check market research results from Chat GPT as it is prone to providing misguided or untrue information. She added that Chat GPT ultimately serves as a springboard rather than a replacement for anything. Even with press releases, it is only an aid as releases still need to have adequate brand voice represented, which can only be properly provided by a human.

“[This period has been about] empowering the teams across our organization to start experimenting and playing around to understand and see the possibilities of where we can trust the machine and where we still need human touch,” she said.

Following this period of AI implementation, Brand Agency London started to turn its attention to using AI in consumer-facing ways. During a Lottie London Halloween campaign, the brand used AI art generator Midjourney to create three art designs. The images were used across the brand’s social media platforms, including Instagram, TikTok and LinkedIn, where it experienced a 710% increase in engagement and a 180% increase in interactions with high social sentiment compared to previous campaigns.

AI also has applications for performance marketing, where Brand Agency London has deployed generative speech AI. The U.K.-based company has found that using local accents in audio ads can improve the engagement of viewers and listeners. Brand Agency London worked with the company Speechify to generate American and Australian accents for its audio ads. The intention is to expand to video in 2024, with a plan to build an in-house branded studio for video production. For example, videos could reflect or represent different markets in the background, while Speechify AI could be used to make an actor or presenter sound native. 

Zukauskaite anticipates that 2024 will be a big year for AI regulation in the European Union and a year for companies to codify frameworks for using AI in interactions with consumers. For example, a brand crediting that Midjourney or ChatGPT was used to generate content can go a long way in informing customers that an image is real or fake. And lastly, as the technology and its users become more advanced, the challenge will be in how brands can use the new capabilities differently from their peers. If everyone uses the same tools in the same way, how do brands continue to differ? Despite generative AI being trained on the entire existence of the internet, it can be rather limited in creativity. For Zukauskaite, this will be the main marketing challenge in 2024.

“That’s where I feel the power is still the creative talent and creative mind of people,” said Zukauskaite. “It’s not that robots will take [someone’s] job, but we will need to find a triangle of how to work between marketers, creative people and technology together.”

Bask:

Miami-inspired sunscreen brand Bask integrated four ChapGPT AI bots into Slack, which assisted with several internal functions. The AI can help develop social media responses and social post copywriting, as well as send customer support responses to inquiries through a customer service portal, and create email marketing copy and imagery. It can also assist with copy for direct mail and other direct response marketing efforts. Lastly, AI helped by serving as a kind of office assistant by using prompts to inform people of how to use the productivity tool Notion, how the company approaches Slack etiquette and when there is an office announcement. Bask worked with Atlas AI, a consultancy and ChatGPT wrapper firm, for implementation.

“Everybody has become more productive. Nobody has to wait for information. There’s a lot of institutional knowledge across any organization, especially at startups, where only one to two people have the answer,” said Mike Huffstetler, founder of Bask. “But now everybody has the answer available to them at all times, 24/7, 365 days a year.”

Based on these positive experiences, Bask created a special ChatGPT chatbot called Bask Pool Boy, which launches in mid-January on Bask’s DTC e-commerce site. Pool Boy can have well-rounded conversations beyond the basic chatbots seen today. For example, Pool Boy can answer questions about when and where to use sunscreen, but it also reveals harbored ambitions of becoming the Bask CEO during customer exchanges. Huffstetler said the idea of building this Microsoft Clippy-like chatbot was to combine engagement with customer service, offering an elevated and humorous take on something functional.

“We don’t want this to be some tool that pushes sales of our product,” said Huffstetler. “We want it to be a value-add to our customers and something they can get lost within. I would love it if 80% of the conversation had nothing to do with sunscreen, and it was all about [people asking] ‘Hey, where are places I should stay and visit if I’m in Miami?’”

Luum Precision Lash:

Since 2017, Luum Precision Lash has been combining robotics and AI to provide eyelash extension services. Unliked Brand Agency London and Bask, Luum relies on predictive AI rather than generative AI, which powers ChatGPT. Predictive AI uses machine learning to identify patterns in past events and make predictions about future events, as opposed to generative AI, which creates novel content, including text, images, videos, code and music.

During a Luum service, a customer is prepped by a technician and then lies down in a booth where two small automated robotic arms work to apply lash extensions. The arms are attached by delicate magnetics, so any light pump will pop them off and prevent injury. Luum has two lash labs — one opened in Oakland, Calif. in 2021 and another that opened in Ulta Beauty in Village Oaks, Calif. in December. Ulta Beauty is also an early investor. While traditional lash extensions with a human technician can take several hours to apply, Luum can complete a full-set lash appointment in 75 minutes. So far, only one eye can be done at a time, but a booth that can complete both eyes is arriving at the end of 2024 and will allow for 33-minute appointments.

The AI comes into play through image processing and decision-making. An example of image processing includes training a neural AI network with thousands of images of eyes to learn where the edge of the upper eyelid is so the robotics knows where to place lashes and where to stop. Examples of AI decision-making include “isolation,” according to Nathan Harding, co-founder of Luum. This refers to the movements of the small robotic arms, with one arm separating a lash from others while the second arm places the extension on the lash with glue. Luum is also working to roll out reinforcement learning in 2024: Once the AI neural network completes the task of informing the robot’s movements, it then determines if the results were good and builds on that information for next time.

“Reinforcement learning can result from techniques such as [lash] isolation [and application] in ways that humans wouldn’t have thought of because we don’t think the same way as a machine,” said Harding.

Harding and Jo Lawson, CEO of Luum, are mindful to point out that the automation of lash extension services does not eliminate the role of technicians. A technician is still required to prep a client and consult them on styles. Harding and Lawson argue that any lash extension business could earn more money by servicing more clients through their devices. How or when AI services could replace human counterparts is an ongoing concern as AI becomes more advanced.

“AI can seem a little scary when you hear things in the news like, ‘AI is going to eat humanity.’ But the way we’re using AI and robotics, so that hardware and the software work together, is proof that AI can make life better,” said Lawson.