Need a little help ordering your next lunch? Starting Friday, fast-casual dining chain Just Salad will roll out a new mobile ordering feature it calls Salad AI, which relies on OpenAIs GPT-4o model to generate text and personalize recommendations that match a customers culinary preferences. Salad AI allows customers to select their dietary restrictions, nutrient priorities, and desired flavor recommendations and within 15 seconds, the AI-powered tool will generate four order ideas.
Its a really powerful tool to navigate hundreds of thousands of potential combinations, says Nick Kenner, founder and CEO at Just Salad, in an interview with Fast Company. We felt super confident that the customer wants help.
Salad AI will be available at all of Just Salads nearly 100 restaurant locations throughout the Northeast, Florida, and the metro Chicago area. The technology will also be added to the web version on desktop computers a little later this year.
The inspiration for Salad AI came about a year ago when Kenner was discussing a personal desire to eat more vegan food, which also incorporated a lot of protein. I said it would be great if AI could tell me what to get on the Just Salad menu, recalls Kenner. And I brought that to our CTO [Matt Silverman] and he said, We can make that happen.
The first of three prompt screens asks customers to select on preferences including vegan, avoid gluten, or minimize carbon footprint. Then, nutrient needs like high protein, low calorie, and low sugar are offered. The final step asks about cravings like sweetness, spice, and savory. In a demonstration, Fast Company opted for a vegan, high fiber, spicy salad bowl and the top recommendation was a spicy vegan fiesta that included kale, almonds, sweet potatoes, corn, and a cilantro lime vinaigrette.
[Photo: Just Salad]
Menu discovery, especially in high SKU restaurant categories like salads, is quite relevant for this [technology], says Mark Abraham, a managing director at Boston Consulting Group who has helped advise restaurant clients deliver more personalized customer experiences.
Abraham says that a large chunk of investments in generative and predictive AI in the restaurant sector tend to focus on the employee and back-office operations. These use cases can make scheduling, hiring, and communication easier for employees, helpful for the industry as it faces a persistent labor shortage. AI can also help chains forecast future demand trends and to manage their inventory.
But the last piece, which isnt as far along, are the customer-facing use cases. Already, predictive AI is being used to help make accurate, personalized recommendations for add-on items in mobile apps popularized by chains like Starbucks and McDonalds. We have seen up to an 8% lift in sales for customers who are getting those personalized offers, says Abraham of the power of predictive AI and machine learning.
Inflation and the pressure on menu prices has led to a dip in demand, with overall spending at restaurants dropping 1.2% between December 2023 and 2024 according to trade group the National Restaurant Association. One way restaurants are looking to become more efficient is through automation, adding more self-checkout kiosks and incorporating robotics to churn out more orders at chains like Sweetgreen.
What it does not replace is human connection, creativity, and innovation, says Fred LeFranc, founder and CEO at restaurant consulting firm Results Thru Strategy.
A few major restaurant chains have honed in on the use of AI-powered voice assistants to take orders from customers at the drive-thru. Wendys in 2023 partnered with Google Cloud to launch FreshAI, tested first in Ohio and since expanded to a few dozen additional locations. But a similar effort at McDonalds, which worked with IBM, led to consumer backlash online because the AI-powered tool was getting orders wrong. McDonalds pulled the plug on the experiment last summer.
Just Salad says it too explored AI in the drive-thru, but wasnt sold on the consumer experience. For now, it is sticking with humans processing those orders, though the chain will continue to evaluate those technologies as they evolve.
Salad AI, Kenner says, also comes with a lot of freedom that allows customers to have the final say on their order. Even after the tool recommends a few customized salads, diners can make modifications before they buy.
Its not intrusive, it is an optional step, says Kenner. It is like a professional assisting you in how to navigate our menu and find what you want.
