Artificial intelligence is the new customer service.
The technology, which has quickly evolved from targeted advertising to “one-on-one” chats with branded bots, can do anything from match foundation shades to helping a customer select a Mother’s Day gift. And brands are counting on these automated, “one-on-one” chats to connect with consumers.
Sephora, L’Oréal, Tommy Hilfiger, eBay, Burberry and Estée Lauder are among a slew of early adopters from the beauty and fashion worlds that are scrambling to deliver personalized experiences online to an omnipresent shopper: a consumer who’s present everywhere from Instagram and YouTube to Snapchat and e-commerce sites, as well as heading to their local specialty retailer or a brand’s freestanding store to interact with their favorite labels. This increasingly savvy customer demands personalization at every step of their shopping journey, which has pushed brands to find innovative ways to integrate artificial intelligence, or AI, into their overarching digital strategies.
In the simplest of terms, AI uses technology to both predict and interact with a customer based on their online habits — and only gets smarter with the collection of more data. For instance, brands and retailers have begun to suggest product recommendations on their digital flagships based on customers’ shopping and browsing patterns, and these suggestions get even more tailored with time. But perhaps the most advanced application for AI to date has been the rollout of Facebook Messenger’s chatbots, which are able to — through automated response — present daily outfit inspiration and help book in-store appointments.
Although AI-based technologies have been around for a few years, they were largely relegated to B2B practices and assisting in targeted marketing programs. This changed last year when Facebook Messenger released bot capabilities in April 2016, giving brands a platform on which to fuel AI-powered communication with customers.
It’s already proving to be an asset for Sephora, which launched the Sephora Reservation Assistant bot via Messenger last fall to help users book makeovers in-store. So far, data shows that booking rates through the bot are 11 percent higher than those that take place on sephora.com.
“We use AI to do natural language processing in the bot to enable it to feel like a chat, [and] it’s actually adding value to our clients….[We’re] able to reduce the number of steps it takes a client to book a makeover,” said Mary Beth Laughton, senior vice president of digital at Sephora, of the simplification of the process, which when facilitated through a bot, took five fewer steps than the traditional booking experience.
While it’s the first true consumer-facing AI experience the retailer has done, Laughton believes a number of applications involving the technology make sense for the brand, especially when it comes to customer service and aiding in a more intelligent search for clients.
“They [Sephora] are leveraging their online channel, but also leveraging their stores — the main asset that they have,” said Anand Chandrasekaran, head of Messenger partnerships at Facebook. He added that brands large and small are building AI experiences on Messenger, which boasts 1.2 billion monthly active users and has seen more than two billion messages sent between people and businesses (a combination of automated responses such as bots and human messaging).
He noted that examples imparting both AI and “rich visual experiences” — such as the Burberry bot for Messenger or Tommy Hilfiger’s TMY.GRL bot that gives behind the scenes content from runway shows and allows users to shop the collection — have resulted in “pretty solid retention.” The experience provides customer service and is fully integrated into tommy.com’s e-commerce site, including stock details. The brand was also the first to launch a conversational, AI bot on the platform.
According to Avery Baker, chief brand officer at Tommy Hilfiger, to date, more than 370,000 messages have been exchanged with an average time of four minutes spent in the bot — two to three times the interaction and engagement time spent on the company’s other digital platforms. About 25 percent are returning users. Additionally, an AI-powered stylist function that kicked off in February at the brand’s Tommyland show quickly became the most popular topic on the chatbot. Forty percent of unique users selected the Style Advice feature, and on average, asked for advice three times or more. The month of February had nearly three times the amount of bot users than the monthly average during 2016, as well as a 71 percent lift of potential revenue for the same period.
“AI-powered conversational commerce is the new frontier in digital and e-commerce. We call it ‘Chat:Shop’ — a new way to bridge engagement with conversion,” Baker said, listing styling advice and increased in-chat content as recent services the brand has tacked on to its chatbot.
Like Sephora, Baker said Tommy Hilfiger is in the midst of “exploring how the platform can be further integrated into other areas of our business.”
On a smaller scale, Epytom, an AI stylist that gives users personalized daily looks based on 40 pieces of clothing they already own, is seeing success with its bot. The two-year-old styling solution sees 100,000 daily users who interact with its bot on Messenger each day.
According to Epytom founder and chief executive officer Anastasia Sartan, 40 percent of users come through word of mouth and 68 percent of users “love” their first look.
“Our engagement depends heavily on whether or not a user liked the first look we send her or him. So we focus on streamlining and shortening our onboarding process while getting as much data about the user as possible, so the very first look we send her is a perfect match,” Sartan said, adding that the goal from the onset was to avoid having users fill out time-consuming forms or photograph their full closets, and instead work to ensure that bot interactions are as seamless as possible.
This is the case at L’Oréal as well, where Lubomira Rochet, chief digital officer of L’Oréal Group, pointed out that it’s easier for customers to express preferences and intent while in the midst of conversation on a messaging app versus having to fill out personal information on a traditional web site.
“The interfaces we use to interact with consumers are going to change tremendously. We’re moving from a web site and up-centric world to a world with more diverse interactions such as voice — we can think of Echo or Alexa — to messaging like WeChat. The windows to the consumer will change and all this is powered by AI,” Rochet said.
AI isn’t new to L’Oréal.
The technology has been part of the brand’s overarching digital strategy for some time via digital marketing platforms such as DoubleClick and Facebook, Rochet maintained, but it wasn’t until April that L’Oréal introduced a consumer-facing AI experience. A pilot of a Beauty Gifter chatbot powered through Facebook Messenger went live on April 19, followed by the launch of personalized video platform Welove.makeup a week later.
Rochet admitted that many competitors launched chatbots following the announcement at Facebook’s annual global developer conference, F8, last year — but her intuition said to hold off for two reasons: the technology wasn’t mature enough at such an early stage and she didn’t want to “build a chatbot for the sake of building a chatbot.”
For her, developing a scenario where AI-based chatbots provided a valuable service for consumers became paramount.
“If you don’t know the utility you want to bring in terms of service it’s not going to help to automate it in terms of AI. If you haven’t spotted something consumers want to do or engage with, there’s only so much AI will help you with. With all marketing, it’s what are the consumers’ needs and how can we use the best technology [to address this] rather than, ‘Cool, there’s a new technology out there and what can we do with it?'”
Maureen Mullen, cofounder and chief strategy officer at digital intelligence firm L2 Inc., agreed, stressing that companies that start with the consumer — versus starting with the technology and developing a use case that provides meaningful value for the end user — will miss the boat. The service-oriented applications for AI lend themselves to fashion and beauty brands (i.e. color matching, one of the biggest obstacles to buying makeup online) and delivering a valuable service to users is the only thing that will actually get people using.
Mullen noted, although the technology is still very much in its infancy, Facebook has had the AI market cornered since announcing the release of chatbots over year ago.
“They [brands] see the investments Facebook has made — not only to Messenger but on Whatsapp and, to a lesser extent, messaging capability on Instagram. Brands see that Facebook’s future is mobile and…there is a lot of desire across innovative organizations to figure out what the application is specifically for their brand,” Mullen said. “It’s smart. If you look at how Gen Z and Millennials are communicating, overwhelmingly, the best customer service you can provide is over a mobile phone. I’d say very few brands have exploited that.”
As far as early adopters go, she cited L’Oréal as “one of the strongest beauty enterprises in terms of pushing the envelope with new technology,” from virtual makeover tools to being among the first to embark on a long-term partnership with a vlogger — Michelle Phan in 2010 — to now trying its hand with emerging technologies such as AI.
The challenge L’Oréal will have going forward, she cautioned, is scaling and ensuring the technology investments it is making — whether it’s AI features or the rollout of new platforms — are more than a novelty or one-off.
For her, the next step is figuring out how AI and chatbots get better integrated into overall communication strategies. The challenge will be the shift from living on a single platform such as Messenger to inclusion within a brand’s e-commerce site or other consumer touchpoints, which could still be three to five years off, Mullen said.
“[They have to] really reach the consumer at scale [in a way] that impacts the business….I don’t think it’s what they’re looking to do right now…[but] kudos for really experimenting, testing and learning….The impetus is now on scale,” Mullen said. “Unless you put a lot of [public relations] and media behind it, most customers never even find it. It’s fun for people in the industry to play with, but it’s more difficult for the consumer to discover in the wild.”
She credited the most advanced AI and chatbot experiences as being done in the travel industry, such as those from Hyatt and Starwood Hotels, which connect clients with the concierge at their hotel or help with booking, respectively.
Stephane de la Faverie, global brand president of Estée Lauder, acknowledged that even though it’s “only the beginning” for AI, results for a Messenger bot dedicated to the brand’s Double Wear foundation are encouraging.
“It’s not only the Millennials. You have everybody interacting,” de la Faverie said of the bot, which works to select a shade and finish with the user and demonstrate how the makeup is intended for use. At the end of the conversation, the individual has a choice to learn more, receive a 10-day sample supply in the mail or purchase the product.
He maintained that click-through to Lauder’s e-commerce site has been strong, and based on early results, the brand will expand the capabilities of the bot to include other hero products.
“It’s a new channel. The same way we would launch a new channel, we’ll find ways to promote it. The consumer is finding a new point of distribution to interact, try on and buy the product,” de la Faverie said. “It’s a full-on integrated experience for the consumer that, in the same way would happen on the counter, happens on your phone or desktop.”