How to identify popular menu items using guest purchase history
Restaurant menus change with health trends, customer tastes, food availability, the season, etc. Here’s how you can use customer data to adjust your menu offerings.
Although you may have an overall idea of what items are popular and what are not, using data to fine-tune your menu can help you avoid food waste, reduce wait times, recommend menu options to guests, and increase your customers’ satisfaction and engagement. But – like so many other things in the modern business world – it all starts with data.
Fortunately, you probably have the data you need, especially if you use an app or website that supports online ordering. Let’s break the process of creating data-driven menu changes into steps and see how we can use this information to shape menu decisions.
3 Steps to identifying popular menu items using data
First, a word of caution: when it comes to collecting personal data from customers, it’s important to get their consent first and to handle that data in compliance with PCI and other regulatory requirements. When customers sign up for online or mobile ordering, obtaining consent is usually part of the process. It’s a good idea to be transparent not only about what data is collected but also how it will be used and who will see it (e.g. “Your data will be aggregated and used by our marketing department to customize offers…”)
Ok, so we’ll assume you have customer order data from your POS, website, and mobile app. Let's see how you can turn that into menu insights.
1: Collecting and storing data
What kinds of data are you collecting? For most quick service and fast casual restaurants, the list includes:
Guest demographic data (age, gender, etc.)
Time/location data
Guest online/in-app behavior (what they’ve looked at)
Time on tasks (e.g. how long they browsed the menu or took to complete an order)
Order/transaction history
Order frequency
Ordered items (including extra items, customized items, and items ordered together)
Device and/or operating system (for online and mobile orders)
This list can potentially go on and on, especially as we have technology that captures these details without any real effort on our part. This leads to the next question:
Where and how is data stored?
This can affect how available data is and how easy (or difficult) it will make the analysis. Aside from the regulatory and security issues around storing guest data mentioned above, it’s important that the data is not siloed – i.e. that data from customer orders can be combined with data from delivery and operations; this allows us to find connections that might not be apparent if we look at each department on its own.
Unfortunately, different tools often store their data in different formats, which makes it hard to analyze data from various sources at the same time. To solve this problem, many restaurants are turning to Customer Data Platforms (CDP). A CDP uses automation to facilitate the collection, storage, processing, and analysis of restaurant data throughout the entire operation. However, we won’t spend a lot of time on that now; we’ll just mention that it’s important whenever you want to do any kind of analysis.
Ensuring data accuracy
Before starting any analysis, it’s important to ensure your data is accurate and relevant. This starts with determining the business goal (in this case, refining the menu) and determining which questions need to be answered. Here, those questions could be “What items have the best/worst cost margins?” “What items should we reintroduce to our seasonal menu?”, etc.
Next, evaluate your data. Look for overlapping or missing information. Make sure the data you’ve chosen will help you find the answers to your business questions. Then check it for completeness; if you’re dividing customers into groups, make sure each group has an adequate number of data points to draw a meaningful conclusion. And also check that you’re not leaving any relevant customer segments out of the equation.
Finally, be on the alert for confirmation bias. We tend to give more weight to results that confirm what we already believe; subconsciously, we may choose data that reflects our beliefs. So try to be as objective as possible.
2: Selecting the right metrics
To choose the right metrics, start at the end and work backward: How will you measure the success of a menu change? Once you know that, you’ll also know what metrics to use in your analysis. Some common choices include:
Order volumes
Store revenue
Food costs
Cost vs. profit margins
Product mix (sold vs. unsold items)
Item popularity
Time to prepare
Customer order frequency
Order type (pickup, delivery, dine-in)
Diner type (solo, family, etc.)
3: Finding trends and patterns in guest purchasing behavior
Deriving information from these data points is all about finding patterns. For example, you may find that certain types of diners order a specific range of foods at a specific time – thus indicating a menu change around certain times of day or week. Or perhaps you see a pattern that some items take longer to prepare, but they are popular during the dinner rush. You can then investigate options to alleviate any service slowdowns, perhaps by featuring faster-prep dishes or prepping more of the popular dish.
Does this sound daunting? Fortunately, a good restaurant analytics tool will do most of the heavy lifting for you. You can also lean on an analytics partner to help you set up the right data and metrics to get the results you need. However you choose to do it, you’ll need to understand how the process works if you want to get the most out of the results of your analysis.
Measuring the impact of menu changes
We’ve discussed how you can choose data and metrics that will uncover your menu’s most popular items. And we’ve talked about finding connections and patterns that tell the full story of why certain items are popular (or not). As we wrap up this article, we still have to discuss one more thing: how to measure the impact of menu changes.
This is relatively straightforward: use purchase and order history information. Simply compare this information pre-change and post-change. Essentially, you’re testing the results of a control group (the old menu) against the test group (the new menu). While you can and should factor in other variables – for example, comparing the Winter 2021 menu to the Winter 2022 menu will usually be preferable to comparing Winter 2022 to Spring 2021, thanks to seasonality changes – simply assessing the new menu or item against its predecessor will give you a good idea of its performance.
In this article, we’ve done a high-level deep dive into restaurant menu analytics. Of course, the specifics of any analysis need to be fine-tuned to your company and goals. In a future article, we’ll zoom out and discuss more ways restaurant analytics can shape your business.
