How to Boost Profits with AI for Restaurants in 2025
By Artificial intelligence is the hot topic in many industries, and hospitality is no different. The problem is that most conversations talk of AI’s potential down the road. What about today? What can we do right now with AI for restaurants as it stands?
Fourth gathered two leaders from Point B, a management consulting firm servicing the hospitality industry, to join Clinton Anderson, CEO of Fourth, for a discussion about the role of AI in 2025. Pat Cormany is a managing consultant of restaurant technology and brings deep knowledge of restaurant operations and opportunities where AI can be applied to simplify business data. He is joined by Lexington Griffith, an AI thought leader at Point B. Griffith works at the intersection of the technology-building team, data and analytics team, and the innovation and design team. Together, these experts provide insights gained from the cutting edge of artificial intelligence and its implementation in the real world.
AI technology can significantly enhance operational efficiency and profitability, so long as it’s done right. In this blog, we’ve assembled the top takeaways from the conversation, including ready-for-action use cases, implementation and adoption strategies, avoiding pitfalls, and maximizing ROI with your AI investment.
Top AI Use Cases Impacting Restaurant Operations
The capabilities of AI could rapidly expand in the coming years. For now, however, there are two key use cases that restaurant operators can act on today that will yield significant cost savings and process optimizations: AI-powered demand forecasting and inventory management.
Use Case 1: AI-Powered Demand Forecasting
Demand forecasting has long been the duty of restaurant operators, and unfortunately, it can be a huge guessing game. Operators glean what demand may look like next month by reviewing past sales records, reservation logs, seasonal patterns, and similar indicators. Over time, an experienced operator can sharpen their instincts and get a sense of the demand, but before that point, there is a lot of trial and error. Each error can mean an overstaffed or understaffed shift, which either inflates labor expenses or damages the customer experience.
One of AI’s most valuable abilities is to quickly analyze large datasets and derive real insights from it. This allows an AI solution to generate highly accurate demand forecasts based on data, rather than intuition. AI-powered forecasting tools gather relevant historical data, weather reports, holiday schedules, community events, and emerging trends to calculate scheduling needs. Moreover, forecast accuracy improves over time as the solution gathers more data. Cormany explains, “You can take some of that forecasting intelligence that an AI engine can generate and continue to learn and improve the accuracy every day and apply that to better labor forecasts.”
AI-powered forecasts also provide valuable guidance in unexpected events. Anderson lays out a compelling scenario in the webinar:
“Imagine you’re operating a restaurant in Portland. It has rained on the same day in April for the last five years. Your forecast is saying, ‘You probably don’t need as many people.’ Out of nowhere, it becomes one of those brilliant spring mornings. There is going to be a warm lunch crowd and another crowd for the warm evening. People are going to be out on the patio.
You know that you need more staff, but you may or may not take that action as a manager. If the AI sees it, it can automatically reach out to team members on the standby list. We know they want to pick up shifts. We know they’ve got the right skills. It sends an invitation saying ‘We’re looking for a bartender, two servers, and one host.’ The AI can receive responses from the team; document their consent, so you’re clear from a compliance perspective; and automatically adjust your forecast for profitability.”
Through the combination of these capabilities, AI-powered forecasting can create significant cost savings. Customers have reported to us 3% savings on total labor spend simply by leveraging AI in forecasting.
Use Case 2: Inventory Management with AI
According to an article published in the International Journal of Applied Management & Technology, up to 10% of the food purchased by a restaurant never reaches the customer. That is a significant amount of waste, but there is a silver lining. It also means one of the largest expenses for the businesses, inventory purchasing, could be reduced by as much as 10% with better inventory management.
Enter artificial intelligence. AI can help restaurant operators secure better prices on ingredients, recognize waste patterns, better distribute inventory purchases across multiple store locations, and anticipate shortages. Each capability helps curb food waste and create savings as a result.
For example, operators can set up auto-ordering policies for ingredients. The AI analyzes purchasing decisions, demand, and shelf-life to calculate the ideal cadence at which to reorder ingredients. Once the AI determines it is time to replenish the inventory, it will review relevant providers to source the ingredient at the best price and purchase a quantity that is projected to meet (not exceed) demand. This allows the restaurant to only acquire the ingredients they need, avoid shortages, and avoid food spoilage.
Blueprint for Smooth AI Implementation
“Where do I start?” It’s the burning question for many operators, but Griffith explained that there are really two questions to consider.
“What are the opportunities? What should I do about them?”
Griffith encouraged interested operators to do some further research into the use cases of today, and if they feel confident about any particular use case, they should treat the deployment as an innovation program instead of a tool implementation.
A tool implementation is a set product purchased for a clear purpose and deployed in the restaurant to meet that end. It’s finite and specific, and much of the process is repeatable from restaurant to restaurant. Innovation programs are different. They are open-ended and exploratory. Griffith explained,
“Whether it’s human processes or just nailing the AI actions to fit the environment, it takes iterations. So, we have to think about it on those terms. Maybe it’s not a six-week program where we know for certain we’re going to get to a level of perfection. Maybe we’re going to assess at three weeks with a lower bar and then reassess in another three and so on.”
Cormany built upon this advice stressing that though there’s an exploratory element to implementing AI, “You have to have those use cases baked in, and you have to have a path to directly achieve results with those use cases to make it a viable improvement to the business.”
It requires a perspective shift from those involved. Executives and other stakeholders should be made aware that implementing AI is not a straight line, it could involve several iterations, and it will not be perfect right away. However, if you can set realistic expectations, secure buy-in, and learn from failures, AI over time will improve operational efficiency.
Once you have approval to move forward with implementation, the best advice is to start small. Rather than deploying to 25 high-value locations to try and maximize ROI right away, find one location that is generally representative of the whole chain. This will allow you to gather insights, refine AI models, and troubleshoot the issues in a manageable environment before expanding.
Empowering Managers to Maximize AI for Restaurants Results
AI evokes a wide range of emotions. Some people are optimistic about a new technological revolution while others fear for their job security. When trying to get managers to adopt AI, it is important to stress that AI is not intended to replace management but rather to support them in their role. Griffith put it this way, “The promise is that with AI, we can actually take some of the unpleasant realities or the small tedious administrative tasks out of their hands.” The benefit of doing so is that AI will free up more time in their day to focus on higher-value activities. We want operators to feel empowered, not threatened, and clear communication is essential in accomplishing that.
Focus on the value to the employee experience. Anderson provided a few compelling examples of messages you can offer your team, such as:
- “It’s going to free up time to do more of the things you want to do.”
- “It’s going to make you spend less time in the back office sitting in front of an applications.”
- “It’s going to allow you to spend more time training your team.”
- “It’s going to allow you to spend more time ensuring a quality guest experience.”
Another element to stress is that whatever AI solution is implemented, it will be straightforward to use. The last thing you want is to provide operators who are already skeptical with an overly-complex interface and a new application to learn. This is a critical step in the implementation process. If the team does not utilize AI effectively, it will diminish the ROI potential of the investment. Cormany put it this way, “You can buy your team a Ferrari, but if you don’t teach them how to drive it…”
AI Is the Game Changer Restaurants Need in 2025
In a business environment with rising food and labor costs, AI offers a way forward that is accessible right now, today. By starting with simple, manageable use cases like labor and inventory forecasting, restaurants can minimize their risk while exploring how best to make use of the technology. It will require some persuading and change management, but the results on the other end include smarter, more efficient operations and increased profitability.
Interested in seeing how AI can help your restaurant? Contact Fourth for personalized guidance on AI solutions and watch the full webinar on-demand.