It’s no secret that AI is here to stay. Of course, the term “AI” has been rebranded before our very eyes whether we realize it or not. Before 2023, people knew that A.I. was actually an acronym for Artificial Intelligence (as we’ve seen in the movie of the same name). But now, if you mention “AI,” people automatically relate it to ChatGPT, Perplexity, Claude, and the literal thousands of other AI tools that have popped up. AI in today’s day and age now means having a digital system do work for you, a shortcut, an automation, and much more.
In the business world, AI can be used for a MULTITUDE of tasks and processes. I’ve seen it used for content, ideation, marketing, automation…a lot of the “tedious” tasks business owners want taken off of their plate. But did you know AI can do SO MUCH more?
While most are still scratching the surface of what AI can do for them, their brand, and their business, AI has been helping clients improve business operations, increase sales, and decrease costs, on top of everything else AI can do.
That’s right, AI can literally act as your own business consultant. If you are comfortable with sharing business information with the tool, it can help you figure out tons about your business:
- Analyze data
- Identify gaps
- Offer solutions relative to your business & industry
- Teach you how to apply ideas that could improve your business
And the best part? You can achieve all of this on most LLMs, like ChatGPT, Perplexity, Claude, on their standard account of $20/mo, where you would probably be paying a business consultant closer to $20 per minute of their time! [Note: An LLM (Large Language Model) is a type of artificial intelligence model designed to process and generate human-like text.]
Allow me to explain. I had the pleasure last year of being asked to keynote the franchisee conference for Copeland’s Restaurant in New Orleans and talk about how AI could impact their business as a whole, as well as how each franchisee could utilize AI to help their business grow. To get me started, they supplied me with two spreadsheets of data: their previous quarter’s labor & product costs, and their sales forecast (which also included the previous quarter’s detailed sales).
I uploaded the documents into Perplexity – my favorite research and internet searching AI tool – and gave it this prompt:
“You are a professional business analyst for dine-in restaurants. Attached is data from Copelands restaurant including labor and product costs & variance. Based off of this data, we are looking for areas for improvement that we can implement to increase sales & margins, and reduce costs where it makes sense to do so. Provide the top 5 areas for improvement based off the data. Ask me any clarifying questions before proceeding.”
This prompt follows Jonathan Mast’s popular prompt building framework which I would highly recommend people use to start their conversation with whichever AI they are using. Here it is broken down in case you need it:
- Give the AI a job/title — Example: You are a professional business analyst for dine-in restaurants.
- Set the scene/describe what you are looking to accomplish — Example: Attached is data from Copelands restaurant including labor and product costs & variance. Based off of this data, we are looking for areas for improvement that we can implement to increase sales & margins, and reduce costs where it makes sense to do so.
- Tell AI what you want it to do — Example: Provide the top 5 areas for improvement based off the data.
- Ask for clarifying questions — Example: Ask me any clarifying questions before proceeding.
Here is what it came back with:
“Before I provide the top 5 areas for improvement, I have a few clarifying questions to ensure I have a comprehensive understanding of the data and can provide the most accurate analysis:
1. What is the target labor cost percentage for Copeland’s restaurant?
2. Are there any specific sales or profit margin goals the restaurant is aiming to achieve?
3. Is there any additional information on food costs or inventory management that’s not included in the provided data?
4. Are there any particular operational challenges or customer feedback that the restaurant is currently facing?
5. What is the current overall financial performance of the restaurant compared to industry benchmarks or previous periods?”
These clarifying questions are ALWAYS important because the output/response AI gives you heavily depends on how clear your information is, as well as how much detail & context you provide it. The more detail & context, the better the response.
Having AI ask you more questions gives you another opportunity to provide more information. If you can’t answer the questions – that’s okay! The best part about AI is that it’s nonjudgmental, because I didn’t know the answers to ANY of the questions! So I literally told it, “I don’t have that info.”
So, it gave me this response:
“Based on the provided data and your additional context, I’ll outline the top 5 areas for improvement to reduce variance in accordance with forecasted sales and potentially increase overall performance
1. Labor Cost Management
The data shows significant variances between scheduled and actual labor hours across various positions. For example
-
AM Server: 660.17 actual hours vs 724.00 scheduled hours (-63.83 hours, -8.82% variance)
-
AM Bar: 148.58 actual hours vs 211.50 scheduled hours (-62.92 hours, -29.75% variance)
Recommendation Implement more accurate scheduling based on historical data and forecasted sales. Use a labor management system to track real-time labor costs against sales and adjust staffing levels dynamically during shifts.
2. Sales Forecasting Accuracy
The overall sales performance is 3% down from forecasted sales, indicating room for Improvement in forecasting accuracy.
Recommendation: Review and refine the sales forecasting method. Consider factors such as historical data, seasonality, local events, and economic indicators. Implement a rolling forecast that’s updated weekly or bi-weekly to improve accuracy.
3. Shift Planning and Productivity
Some positions show significant over-staffing or under-staffing
-
AM Utility: 337.93 actual hours vs 190.00 scheduled hours (+147.93 hours,+77.86% variance)
-
AM Line: 868.78 actual hours vs 714.00 scheduled hours (+154.78 hours, +21.68% variance)
Recommendations: Analyze peak business hours and adjust shift start/end times accordingly. Cross-train employees to handle multiple roles during slower periods, allowing for more flexible staffing.
4. Overtime Management
Several positions show overtime hours, which can significantly impact labor costs:
-
AM Line 13.93 overtime hours
-
AM Prep: 19:90 overtime hours
Recommendation: Implement stricter overtime policies and approval processes. Use scheduling software to alert managers when employees are approaching overtime thresholds.
5. Menu Engineering and Product Mix
While specific food cost data wasn’t provided, optimizing the menu can help improve sales and reduce variances.
Recommendation: Conduct a menu engineering analysis to identify high-profit, high- popularity items. Promote these items and consider removing or redesigning low-performing dishes. This can help increase average check size and potentially reduce labor needs in food preparation.”
Remember, this took less than 5 minutes to get this far. I uploaded the documents, gave it my prompt, the AI asked me clarifying questions, and as I didn’t have additional information to provide, it gave me this response.
What would take a professional HOURS to produce this list of 5 items, took AI less than 5 minutes.
Reviewing the list AI gave me, I noticed that three of the five items focused on the AM shift work. After taking some time to think about what direction I wanted to take this conversation with AI, I had an idea to give it a follow-up task:
“AM shift planning appears to attribute to increased costs. Create a scheduling plan that lowers variance while ensuring we are properly staffed for all shifts.”
A short, simple task. I didn’t even give it much detail, I just basically said, “Hey, I noticed this, can you give me ideas on how we can fix this common denominator.” Here’s what it came back with:
“Based on the analysis of the labor data provided, I’ll outline a scheduling plan to lower variance while ensuring proper staffing for all AM shifts. This plan aims to optimize labor costs and improve efficiency.
1. Implement Data-Driven Scheduling
Use historical sales data and foot traffic patterns to create more accurate forecasts for each day of the week. This will help determine the optimal number of staff needed for each shift.
-
Utilize the scheduling software’s forecasting features to predict busy periods
-
Analyze peak hours and adjust staff levels accordingly
-
Create shift templates based on typical business patterns (e.g., weekday lunch, weekend brunch
2. Staggered Start times
Implement staggered start times to better match staffing levels with customer demand throughout the day
Example:
-
6:00 AM: Opening staff (prep, kitchen)
-
7:00 AM First wave of servers and hosts
-
9:00 AM: Second wave of servers for late morning rush
-
11:00 AM: Additional staff for lunch service
3. Cross-Training Program
Develop a cross-training program to create a more flexible workforce. This will allow you to adjust staffing levels more easily during unexpected fluctuations in business.
-
Train servers to handle hosting duties
4. Optimize Shift Lengths
Adjust shift lengths to reduce overtime and better match labor to demand.
-
Use 6-hour shifts for peak periods instead of full 8-hour shifts
-
Implement 4-hour shifts for specific roles during predictable rush periods
5. Establish a Core Team and Flex team
Create a core team of full-time employees supplemented by a flex team of part-time staff
-
Core team: Scheduled for consistent shifts based on their availability
-
Flex team: On-call for busier days or to cover unexpected absences
6. Implement a Time-Off Request System
Use the scheduling software’s time-off request feature to better manage employee availability and reduce last-minute schedule changes.
-
Require time-off requests to be submitted at least two weeks in advance
-
Use the system to automatically block off approved time-off when creating schedules
7. Create a Shift Swap Protocol
Establish a clear process for shift swaps to give employees more control over their schedules while ensuring proper coverage.
-
Use the scheduling software’s shift swap feature
-
Require manager approval for all swaps
-
Set a deadline for swap requests (e.g., 24 hours before the shift)”
Whoa. Within five minutes, I had seven actionable ideas I could bring to the restaurant owners. Notice I said ideas and not solutions. AI is NOT perfect. It doesn’t have all of the answers. There is no guarantee that if the restaurant applied all seven of these that they would see an immediate impact. AI can not account for human intervention and behavior. Think of it like learning in school — the textbook is going to give you examples and situations where, in a perfect world, everything would work as expected…but we’re human and nothing is perfect.
Having said that, I can give these ideas to the client and tell them that these are ideas they can take to their team and then decide which ones (if any) would make sense for them to implement. And again, this took me all of maybe 10 minutes of actual work to get to this point.
As a business owner, you can do the same! If you’ve been keeping up on your data — revenue, costs, etc. — you, too, can upload the information into AI and tell it to analyze YOUR data and give you ideas on how you can improve your operations, your processes, etc. What is even better is that you get to talk to the AI about your data. Think of AI as your own consultant with a wealth of knowledge — you just have to ask the right questions to get the information you are looking for!
AI isn’t just some futuristic sci-fi concept—it’s here, it’s now, and it’s already changing the game for business owners who know how to use it. And the best part? It’s not just for big corporations with deep pockets. With a simple prompt and the right tool, you can have AI analyzing your data, identifying opportunities, and giving you ideas that could take your business to the next level—all in minutes, not hours.
But here’s the thing: AI doesn’t replace strategy, decision-making, or good ol’ business instincts. It’s not a magic wand, but it is a serious shortcut. It takes what used to be time-consuming (or expensive) and makes it faster, easier, and way more accessible.
So, if you’ve been sleeping on AI, consider this your wake-up call. Your business already has the data—AI just helps you make sense of it. The only question is: Are you going to use it to your advantage, or let your competitors figure it out first?
