SAN FRANCISCO — No one knows customers better than a bartender — except AI.
"From an experiential perspective, what would make someone buy a cocktail?" What about two cocktails, Sherif Mityas, chief experience officer at TGI Fridays, asked attendees at The AI Summit in San Francisco Wednesday.
As a social brand, TGI Fridays is all about the customer experience. The goal is to sell and engage more with the patrons that come into its 900 restaurants worldwide.
The restaurant chain is one of many brands that has taken to technology to engage with customers. The overarching goal is to use technology and data to create a better and more tailored experience for the consumer. Artificial intelligence has become part of that.
But "when we think about AI, it's not about the technology," Mityas said. His job is to "worry about the guest."
Understanding what the guest wants requires data. And in a smartphone-enriched culture through mobile platforms, Fridays is able to capture what customers are ordering, when, where they work and who they're with. From that, Fridays can decide when to contact customers and on which platform.
When Fridays first started directly targeting its audience, it did not understand guests, Mityas said. Thousands of emails were sent out without customer differentiation, which meant an email about pork ribs could go out to a vegetarian.
By integrating AI on platforms and through email campaigns, Fridays was able to change that. Through bots on Facebook, Twitter, Instagram and Alexa, the company increased social engagement 500%, according to Mityas.
AI efforts also increased online ordering, reservations and social engagement while in restaurants, resulting in 100% increase in online revenue growth.
Successful case studies created the business case and performance metrics for additional AI implementations. TGI Fridays was "not putting AI in for the sake of AI," Mityas said. Instead it was driving a real business performance result.
The key is "there is no reward for speed here," Mityas said. With AI, "you've got to trial and error. You've got to make time for you, the team and the tool to learn. You have to be in lockstep. It has to be together."
The case for cocktails
TGI Fridays worked to create the feel of a neighborhood bar, where the servers knew customers' usual and the bartenders knew their name. It focused in particular on cocktails, a high-margin item.
"All those $10 drinks you guys buy out there do not actually cost $10," Mityas said. It's good for business if customers buy one cocktail. But if they purchase multiple, profits increase even more.
Using AI, Fridays worked to increase the intimacy of the guest-bartender relationship without tarnishing the warmth of the interaction.
Along the way, Fridays had to solve a couple of problems:
-
The restaurant chain had to operationalize the technology in a sensible way. In this case, it was implementing an AI tool that could understand every drink ever made. The availability of the technology made it easier.
-
The company also had to determine logistics for implementation. Bartenders had to be able to craft a wide array of drinks. "I had to buy all these restaurants a new spice rack, because believe it or not there's some crazy spices that go into some of these drinks," Mityas said.
Fridays created a personalized cocktail platform on the TGIF mobile app and using questions the virtual bartender "Flanagan" could help the customer craft a beverage, sending the end result to the bartender.
Flanagan would ask questions like, "why are you here" or "what are you going to be eating" to help a guest create a cocktail to fit their immediate mood, Mityas said. Numerous choices, flavors and tastes allowed customers to create unique mixes and at the end, they got to name their drink (through prompts, not open text boxes, which could solicit unsightly responses).
When the AI tool does its job, it creates a one-of-a-kind drink just for that guest at the bar. There's "no perfect, better use case than selling alcohol with AI," he said.
Through the technology, Fridays was able to increase the number of people who bought two drinks by 25%.