How to Get More Guests With Service & Hospitality Using AI
Jon Asher isn’t shy about admitting how quickly the ground has shifted beneath restaurant technology leaders. As chief technology officer at Nékter Juice Bar, the fast-growing juice and smoothie chain with more than 200 locations and roughly three million loyalty members, Asher joined Fast Casual Nation hosts Paul Barron and Cherryh Cansler to talk about what artificial intelligence actually looks like on the ground at a scaling restaurant brand. His answer wasn’t about flashy chatbots or gimmicks. It was about data. Nékter currently operates across three point-of-sale systems as it consolidates toward one, a process Asher said is essential to building the kind of clean, unified data foundation that makes AI tools worth using in the first place.
That foundation, Asher explained, is the real dividing line in how AI performs for restaurant brands. He described the difference between “rented” data, where third-party vendors manage and often anonymize customer information inside their own platforms, and first-party data, where a brand builds its own warehouse and controls how its point-of-sale, loyalty and review data connect to one another. Nékter uses a mix of both, but Asher was clear that the real value comes from linking those data sets together so AI tools can ask sharper questions, like predicting which guests are at risk of churning based on order history and review sentiment.
Personalization was where Asher got specific. Nékter has moved well past blanket “we miss you” campaigns triggered after 45 days of inactivity. Today, the brand’s AI-driven approach can distinguish between a lapsed regular who needs a stronger incentive to return and a family that simply went on vacation and will likely come back without any discount at all. Asher said the company measures the impact using holdback groups, essentially a control group of guests who receive no offer, to isolate how much incremental revenue AI-driven campaigns actually generate.
Asher was equally candid about where AI’s limits show up. He compared frontier AI models to “an extremely smart toddler” with the credentials of an Ivy League graduate but none of an operator’s real-world experience, meaning human oversight and dashboards remain non-negotiable. He also drew a firm line around hospitality itself: while AI has improved things like personalized upsell recommendations and customer service email response times, Asher said he still believes there’s a point where guests want a human to step in, particularly when something goes wrong with an order.
On vendor trust, Asher said Nékter is willing to hand sensitive data to specialized providers, like image generation tools, but keeps guest and loyalty data close, having been burned before by vendors whose privacy promises didn’t hold up. He noted the team currently leans on Claude for coding-related work given its existing infrastructure, though he acknowledged the competitive landscape between AI providers shifts month to month. For loyalty growth, Asher pointed to phone-number-based redemption, which replaced a glitchy QR code system about a year and a half ago, as one of the simplest but most effective changes Nékter has made to reduce friction at the register.
Asher’s closing advice to fellow restaurant technology leaders was less about chasing the newest AI tool and more about discipline: get the data clean, get it connected, and start now, since he believes the brands building strong first-party data foundations today will be the ones best positioned to scale personalization years down the road. Barron and Cansler are continuing their series on AI in restaurant operations on Fast Casual Nation, with more guest interviews planned ahead of the Fast Casual Executive Summit, October 4-6 in Arlington, Texas.
