How restaurant marketing can use identity management and machine learning to drive one more visit and one more dollar.
February 22, 2019
Sponsored Content
Restaurant operators understand that personalization is a crucial component of their marketing strategy. In fact, recent research on restaurant marketing from Nation’s Restaurant News and Epsilon-Conversant shows that 70 percent of operators say personalization is either very or extremely important in their marketing messages.
However, the ability to deliver relevant, personalized communications—messages that truly resonate with each person—is completely dependent on the quality of the marketing technology being used and the data inputs to the system.
For restaurants, this becomes even more complex with information coming in from different sources and streams across physical, digital, email, CRM, loyalty and online ordering platforms. Many restaurant marketers struggle to understand how to use all the data that is available and to make it actionable for their marketing plans.
In many ways, it comes down to effective identity resolution—the ability to know and recognize the same person across all touchpoints—and machine learning. Paired together, knowing people (and their interactions with a brand) and then using machine learning to recognize patterns in their behavior informs what marketing messages each person should receive next, through which channel, on which device and at what time of day. When done well, these two components allow brands to accurately identify and interact with their audiences at any touchpoint—from email and loyalty apps to digital display ads and—the end goal—in-restaurant orders.
Here, let’s take a look at the current challenges restaurants face in these areas and offer recommendations to leverage both concepts for more effective, holistic restaurant marketing efforts.
Inhibitors to adoption
In the research, “Driving one more visit: How restaurant marketers fare in the digital age,” a few nuances about how restaurants currently use marketing channels became clear:
In addition to ordering at their physical location, roughly 50 percent of restaurants allow customers to place orders through different channels, such as third-party delivery services, the restaurant’s website or their mobile app. This is up to four different options for consumers to purchase or interact, which can cause more disconnect between knowledge about a single customer.
Operators are good at personalizing based on a customer’s physical location, but less adept at more complex variations like being able to personalize based on a combination of channels, messages, offers and message frequency that a customer is most likely to respond to.
Only 29 percent of respondents are very confident that their CRM and technology infrastructure allows them to activate cross-channel customer data to fuel their marketing programs. That points to missed opportunities as operators are spending resources on technology and not using the platforms to activate personalized marketing communications, and Epsilon research on personalization shows that 90 percent of consumers are more likely to do business with a company that offers personalized experiences.
Only 49 percent of respondents say they use loyalty program information to understand/analyze customer value. And only 30 percent of respondents with loyalty programs say their programs are effective.
These findings suggest that restaurant operators have some work to do with understanding their customers holistically and then delivering relevant, personalized messages to them in the moments that matter.
Bringing disparate streams together
Restaurants are new to the data-driven marketing game. Retail brands have been doing this for years through online orders paired with in-store transactions that allow them to understand their customers holistically. Restaurants, on the other hand, are more recently activating cross-channel ordering with robust loyalty programs and need ways to bring that information together with offline transaction data to prove their efforts are actually working.
A great example of understanding this single customer view is Dunkin’ Donuts. In 2013, they started their DD Perks rewards program; today, it has more than 6 million active, loyal members. Why is it successful? Dunkin’ customers are incentivized to use the DD Perks app for easy, convenient payment and to earn free coffees. Meanwhile, the chain is able to send customers digital messages and discounts through the app and connect them back to the POS. By knowing the individual and their unique purchase habits, Dunkin’ uses machine learning to manage and analyze the large pool of customer data and create customized content for DD Perks members. Instead of a generic offer such as getting the 10th coffee free, guests receive personalized messages and offers based on their preferences and purchase habits.
Average year-over-year spend for new DD Perks members increased 40 percent, and member discounts are driving both increased visits and higher average weekly sales. This shows that pairing that single view of a customer together with machine learning to help determine how and when to interact with each person can drive real results for brands at scale.
Recommendations to deliver true personalization
For foodservice operators, the key to driving one more visit—one more dollar—from current customers is to start delivering truly personalized interactions in every communication.
Start by organizing first-party data in a way that shows each contact as a real person, across all of their activity with the brand.
Then, pair CRM information with third-party transaction, historical and spend data to understand customers through more than just a single brand’s lens.
Next comes the data activation. With one holistic view of each person across interaction points, restaurants can start using machine learning (through marketing technology partners) to make decisions about what messages to send, what time, what content, etc. in order to optimize every interaction.
Together, this combination of data streams and sources creates a singular view of each individual, allowing brands to deliver personalized messages, and make the most out of every customer touchpoint. And—ultimately—this method allows the brand to see whether or not the marketing actually influenced that person to visit the restaurant one more time.
Interested in learning more about the convergence of identity and machine learning? Go to epsilonconversant.com for more information.
