ChatGPT recently released a new feature called Code Interpreter, that allows you to upload a spreadsheet of data and ask for an analysis. So I asked it to ‘Give me 10 ways ChatGPT’s new Code Interpreter could analyse and improve menu profitability for a restaurant or cafe? Give answers that appeal to a restaurateur or accountant with fairly sophisticated financial understanding and good analytical skills.’ Here’s what came back:
ChatGPT’s Code Interpreter can add significant value by analysing and improving menu profitability in the following ways:
1. Menu Engineering: The Interpreter could analyse each item’s popularity and profit margins to classify them into four categories – Stars, Plow Horses, Puzzles, and Dogs. This can help identify items that should be promoted or repriced, based on their popularity and profitability.
2. Price Elasticity Analysis: By studying historical sales data and price changes, the Interpreter could estimate the price elasticity of each menu item, helping identify items where a price increase might have a minimal impact on sales volume, and thus potentially increase profitability.
3. Contribution Margin Analysis: By determining each item’s contribution margin (selling price less variable cost), the Interpreter can help identify which items bring the most marginal profit per sale. This could influence decisions about pricing, promotion, or even menu design.
4. Cross-Elasticity Analysis: The Interpreter could analyse sales data to determine how changes in the price or availability of one item affect the sales of other items. This information could be used to manage complementary or substitution relationships between items to increase overall profitability.
5. Seasonality Analysis: By comparing sales data from different periods, the Interpreter can identify seasonality patterns in sales. This could inform decisions about rotating menu items or running promotions to maximise profitability at different times of the year.
6. Waste Analysis: By comparing inventory and sales data, the Interpreter can identify areas of waste – items that are often thrown away or ingredients that spoil before use. This could drive changes in portion sizes, ordering practices, or menu design to minimise waste.
7. Ingredient Cost Analysis: The Interpreter can analyse ingredient costs and sales data to identify potential savings. For instance, if a popular, high-margin item and a less popular, low-margin item share a costly ingredient, dropping the less popular item might be worthwhile to save on ingredient costs.
8. Customer Preference Tracking: The Interpreter could be used to analyse customer feedback data to determine preferences and dislikes, informing potential changes to the menu or ingredients to better appeal to customers and drive sales.
9. Profitable Pairing: The system can identify commonly ordered pairs or sets of items by analysing transaction data. Creating promotions or deals around these combinations can enhance customer value perception and boost overall profitability.
10. Competitor Pricing Analysis: The Code Interpreter could be programmed to analyse publicly available menu data from competitor restaurants. By comparing prices for similar items, you can gain insights into the market standard and adjust your own pricing accordingly.
I was intrigued by #9 – Profitable Pairing, and asked for more information…
“Profitable Pairing” involves analysing the restaurant’s transaction data to find patterns of frequently ordered items. This method is similar to what is often called “market basket analysis” in retail settings. In the context of a restaurant, the “basket” would be a single order, and the items in it would be the dishes and drinks ordered by a customer or a group of customers.
When you have enough data, you can start to see trends. For example, you might find that a certain wine is often ordered with a specific dish or that a particular appetiser is popular with customers who order a certain dessert.
Once you’ve identified these patterns, you can use them to create special offers or promotions. For example, if you notice that a particular meal and dessert are often ordered together, you might bundle them together as a “special” and offer a small discount. This can increase the offer’s perceived value, encouraging customers to spend more than they might have otherwise.
Another benefit of profitable pairing is encouraging customers to try new items. If a popular dish is often paired with a less-known item, bundling them together can encourage customers to try something new, potentially increasing the popularity and sales of the less-known item.
This kind of analysis can also influence menu layout. For example, if you know a certain starter is often ordered with a specific main, you could place these items near each other on the menu to encourage customers to order the pairing. You can also optimise inventory management by identifying items that often sell together. For instance, if a dish that uses a lot of a certain ingredient often sells with another that uses very little of the same ingredient, you can reduce waste by carefully managing that ingredient’s stock based on the popular pairing sales trend.
I also asked for more information on #2 – Price Elasticity:
Price elasticity of demand is a measure of the relationship between a change in the quantity demanded of a particular good and a change in its price. It is a crucial concept in economics and particularly useful for a restaurant to understand how price changes might affect sales volume. When the price of a menu item goes up, customers may buy less of it, and when the price goes down, they may buy more. However, the extent to which this happens varies from product to product, and that’s where price elasticity comes into play.
This should be much better appreciated by people who argue that the quality of their ingredients means a high price is ‘justified’… maybe, but you will sell fewer of them!
If a product is said to have high elasticity, it means the quantity demanded is quite sensitive to changes in price. For example, if a 1% increase in the price of a dish leads to a 1.5% decrease in quantity sold, the price elasticity of demand would be -1.5, indicating a highly elastic demand. In contrast, if a 1% increase in price only leads to a 0.5% decrease in quantity sold, the price elasticity of demand would be -0.5, indicating inelastic demand.
Code Interpreter could analyse historical sales data and menu price changes to calculate each dish’s price elasticity. With these insights, you can make more informed decisions about your pricing strategy. For example, you could increase the price of items with inelastic demand to improve profitability since a price increase won’t significantly impact the quantity sold. Conversely, for items with elastic demand, you can keep prices steady or even lower them to increase sales volume and overall revenue.
However, this approach has certain limits: changes in quantity demanded could be due to other factors like changes in consumer income, tastes, or the prices of other goods. Isolating the impact of price changes can be challenging. You also need good historical data, and there may be a non-linear relationship between price and quantity demanded, especially over large price changes. You also have to deal with customer and competitor reactions.
I asked ChatGPT about finding profit opportunities on a wine list:
Market Segmentation Analysis: The Code Interpreter can help identify key market segments that patronise your restaurant by analysing sales data. For example, you may discover that business customers prefer certain wines while others are more popular with couples or groups – this can guide the pricing and promotion strategy for each segment.
Promotion Effectiveness Analysis: If you run wine promotions, the Interpreter can analyse their impact on sales and profitability. This can include “happy hour” promotions, discounts on certain days of the week, or deals when paired with certain meals.
Dynamic Pricing: Code Interpreter could use historical data to develop a dynamic pricing model for wines, where prices are adjusted based on demand patterns, time of the day, day of the week, or season. For instance, if certain wines sell better on weekends, you might increase their price slightly while reducing it on weekdays to encourage sales.
Purchase Frequency Analysis: Code Interpreter could identify wines that turn tables quicker (i.e., consumed faster), driving higher total sales. These wines might be candidates for a small price increase.
Stock and Sales Correlation: By correlating inventory levels and wine sales, the Code Interpreter can help determine if there’s a point where the lack of variety affects overall sales. This can influence decisions about how much variety to offer and how to price to reduce inventory while maintaining sales.
Customer Feedback Analysis: By integrating with a system that collects customer feedback, the Interpreter can identify trends or issues with particular wines (e.g., consistently rated as poor value for money). This could inform decisions about which wines to keep, replace, or adjust pricing on.
Price Bracket Analysis: Code Interpreter could analyse sales data to determine the most popular price brackets for wine in your restaurant. This can help you identify gaps or opportunities in your wine list – for example, if you have a lot of wines in the $30-40 range but most sales are in the $50-70 range, you might want to source more wines in the latter range.
Have fun using ChatGPT’s Code Interpreter to drive up your restaurant profits – it’s all about good prompts and good data for it to work with!
See also: How to Use ChatGPT for Recipe Management in a Restaurant Kitchen and also How to Use ChatGPT to Create Training Material for Restaurant & Cafe Employees.
🤚 Check the weekly discoveries on Hospo Reset – information & inspiration for restaurant, cafe & foodservice operators.