Restaurant technology trends to watch out for in 2020

According to multiple outlets, 2019 was the year of the chicken sandwich. Social media feuds and new product debuts kept fried chicken top-of-mind within the QSR industry and beyond. But the chicken sandwich wasn’t the only important development of 2019. In our view, the theme of the year was lasting developments in technology: restaurants bet on big data and modified their operations to accommodate the continued rise of digital ordering. However, it’s clear that the movement has just begun. Here, we take a look at the trends that defined 2019 and are expected to go strong well into 2020. 

Betting big on technology

In the National Restaurant Association’s 2019 State of the Restaurant report, 70% of quick service restaurant operators planned to devote more resources to investing in technology, with 57% planning on investing more in back-of-house operations like POS systems and 41% planning on investing more in customer-facing technology like self-order kiosks. 

In collaboration with Baidu, China’s largest search engine, KFC debuted a facial recognition system designed to predict personalized menu options based on a customer’s age, gender, and mood. In collaboration with Yext, Taco Bell is enhancing its digital presence, ensuring that the brand pops up when customers use search terms like “fast food”, “Mexican food”, and “drive-thru.” 

Large multinational companies who are particularly well-placed to make investments in technology are at the forefront of this movement. McDonald’s acquired Dynamic Yield, a startup that provides retailers with decision logic technology, for $300 million in March. Just a month later, McDonald’s also announced it had acquired a 9.9% minority stake in Plexure, a mobile-app vendor. Then, in September, McDonald’s acquired Apprente, a company that builds conversational agents focused on fast-food ordering. Many predict that this will be another big year for M&A in the restaurant space. The year already kicked off with big news about the Yum! Brands acquisition of Habit Burger. But in 2020, will more restaurants follow McDonald’s lead in acquiring tech companies?

Using artificial intelligence in novel ways

McDonald’s acquisitions of Dynamic Yield and Apprente in 2019 reflected the industry trend of investing in artificial intelligence. With both of these acquisitions, McDonald’s is aiming to overhaul the entire drive-thru experience, using Dynamic Yield’s technology to show customers personalized drive-thru menus and Apprente’s technology to automate voice ordering. 

Other companies are making more modular changes. To improve phone ordering, Chipotle is testing a conversational voice bot, using voice recognition technology to interpret customers as well as machine learning to improve the algorithm after every conversation. In the field of marketing and rewards, TGI Fridays is using artificial intelligence to personalize mobile device notifications, and Punchh recently closed a $41 million round of funding in order to augment its AI algorithms, which generate targeted multi-channel marketing campaigns in order to foster brand loyalty. Other fun applications include Domino’s DOM Pizza Checker, which uses computer vision to check the quality of pizzas before they are delivered to customers in Australia and New Zealand, and Chick-fil-A’s customized artificial intelligence system that predicts foodborne illnesses based on social media mentions. 

Redesigning stores to be digital-first

To grapple with the ever-increasing influence of digital ordering, companies are testing out new store formats that prioritize off-premise orders. 

Some of these changes have been incremental, building infrastructure for customers to pick up their digital orders in a designated space. Firehouse Subs, a Florida-based sandwich chain, debuted a new restaurant format with an emphasis on pickup shelves to accommodate the off-premise orders that now make up 62% of sales. In a variation of the Firehouse pickup shelves, Pizza Hut debuted a new location with carry-out pizza lockers. 

However, some restaurants are also piloting takeout-only models. Marking the first time the international burger chain has debuted a new store format since launching drive thru restaurants in the 1980s, McDonald’s unveiled a new, take-out only store in London with kiosks and a reduced menu for convenience. Similarly, KFC opened an experimental drive-thru only location in Newcastle, New South Wales in November. Even delivery services like DoorDash are getting in on this trend: DoorDash is piloting a series of “ghost kitchens,” offering dedicated restaurant space to some of its partners (The Halal Guys, Nation’s Giant Hamburgers, Rooster & Rice, and Humphry Slocombe have already signed on) to prepare orders exclusively for DoorDash deliveries. We’ll see interesting results in 2020 as these innovative new restaurant models are launched and evaluated. 

The rise of the self-order kiosk

Major players like Burger King and McDonald’s have been experimenting with kiosk ordering since the mid-2000s, but it was not until recently that the restaurant industry began to see the mass adoption of self-order kiosks. This shift was set into motion by major restaurant brands like Subway, Panera, and Wendy’s, who began testing kiosks in between 2015-2017. Today, self-order kiosks are in two-thirds of Wendy’s locations. 

However, kiosks are no longer only reserved for major restaurant brands who can afford to make large technological investments. Companies like Bite, offering out-of-the-box solutions to smaller restaurant brands, are democratizing this valuable technology and making it available to brands of all sizes. In 2020, we’ll expect to see even more restaurants jump on the bandwagon with kiosk ordering, especially as it becomes an operational model that customers have grown to expect.

Unlocking the power of machine learning with a turnkey solution

By offering an integrated solution for companies both large and small to leverage the power of machine learning in their operations, Bite Kiosk works with a variety of restaurants across the industry, allowing restaurants to focus on what they do best: providing great service and making great food. Bite’s algorithm uses reinforcement learning to provide recommendations to customers, and Bite Kiosk works out of the box to integrate seamlessly into existing workflows and deliver results. 

Steve Truong, Head of Product at Bite, explains how machine learning algorithms predict recommendations by extracting information from the existing data they’re provided. “Machine learning algorithms take a large dataset and use the data to train the algorithm. The algorithm’s parameters determine the output given some input,” says Truong. The algorithm’s parameters could include the preferences of other customers at the restaurant, environmental factors such as weather and time of day, and the user’s own historical preferences, if they’ve previously opted into facial recognition. 

In particular, the Bite algorithm is a reinforcement learning algorithm that improves over time. Truong explains, “Over time, you reinforce the learning by feeding the algorithm data and telling the algorithm whether or not the answer it predicted was right or wrong. Then, it will internally adjust its own parameters to more closely approximate a better guess the next time it gets asked the same question. As you feed the algorithm more and more of this data, it will get better and better at seeing patterns and produce more accurate results.” 

Because of the machine learning algorithm, Bite’s recommendations are more powerful and flexible than upsell recommendations from cashiers. “Just saying ‘Would you like fries with that?’ is not enough. The machine learning we’re developing asks Would you like ‘X’ with that?” where ‘X’ is what the customer would say yes to every time,” says Truong. 

Ultimately, Bite Kiosk is a turnkey solution that allows companies to bring the power of machine learning to the guest experience without additionally complicating the restaurant’s operations. “Our solution comes in an integrated package. We’re not a general purpose machine learning consultancy,” says Truong. 

Unlike data analytics derived from a machine learning consultancy that have to be consciously integrated into the workflow, Bite Kiosk applies its findings immediately. “Using machine learning to improve restaurant operations doesn’t make sense until you can put it into practice, and kiosks are the ideal way to put it into practice. Kiosks are the perfect medium for reinforcement learning because they provide instant feedback. As we learn your taste profiles, we can reorder the entire menu board on the kiosk based on what you like.” Bite gives its clients the ability to provide customers with entirely bespoke guest experiences tailored to their individual needs and preferences. In doing so, Bite is leveraging technology to revolutionize the restaurant guest experience, enabling restaurant brands to provide their guests a level of personalized hospitality never before possible.

How leading restaurant brands are driving productivity with machine learning

Machine learning (ML), a subset of artificial intelligence, refers to systems that can learn from experience to solve specific tasks. ML algorithms use statistics and linear algebra to find and apply patterns in massive amounts of data that may be invisible to the naked eye. In recent years, ML algorithms have transformed consumer products ranging from music and video streaming services to search engines to retail e-commerce. By giving consumers personalized recommendations derived from analyzing the data they provide, these companies can optimize their products for each user, in essence providing a bespoke product to every single one of them. 

While the restaurant industry has been slower to adopt machine learning technology than other industries like those mentioned above, innovators at the biggest companies in the restaurant space are increasingly seeing that their firms cannot afford to ignore these technologies. Top executives at Domino’s even refer to the company as “an e-commerce company that sells pizza,” emphasizing the company’s tech-driven mindset. Below are some notable examples of restaurants driving productivity with machine learning:

Customer Experience

McDonald’s reached an $300 million acquisition agreement with Dynamic Yield, a machine learning startup that works with brands to personalize customer experiences. In particular, McDonald’s acknowledged that it planned to use Dynamic Yield’s decision technology to present customers with intelligent, dynamic Drive Thru menu displays based on factors like trending menu items, weather, and time of day. 

Food Safety

Chick-fil-A developed custom technology to track social media mentions in order to identify outbreaks of foodborne illnesses. Using custom software, Chick-fil-A is able to predict the likelihood of an emergent illness by identifying trends in phrases and keywords used in social media posts related to the brand.

Marketing

Dunkin’ Donuts is working with big data firm Splunk to drive loyalty through targeted promotions. Splunk provides insights into Dunkin’s customers’ habits and preferences, allowing Dunkin’ to target specific customers with offers and promotions relevant to them. 

Quality Assurance

Domino’s recently launched the DOM Pizza Checker in Australia and New Zealand, which uses computer vision to check the quality and consistency of pizza pies before they are delivered to customers. 

———

It’s easy to see, however, that these decisions were made by large multinational companies who were able to put up large upfront investments in order to leverage machine learning solutions. Having identified that not every restaurant can afford to acquire technology startups, contract with consulting firms, and/or build technical solutions in house, some emerging restaurant tech companies are making this technology available to the industry at large. 

Read on to the next article to learn about how Bite is using machine learning to revolutionize the in-store ordering experience for restaurant brands both large and small.

Read this before you decide to build a self-order solution in-house

Many underestimate the commitment that goes into owning and maintaining a piece of software. We’ve put together a guide that explores some of the underestimated costs of building software, and why it likely makes sense to look for a preexisting SaaS solution.

Last month, Stripe engineer and tech writer Noah Pepper published an article titled “The Treacherously High Cost of Software”. The article touches on a question that many decision-makers in our space face regularly: when to find a preexisting software solution, and when to build it in-house? As Noah’s article illustrates, the answer is that in most cases building software in-house should be avoided. That is because it is far more expensive than many people realize.

So why is software so expensive to maintain?

The world is always changing.

As Noah writes, “Just treading water costs money if you want to be compliant, secure and operable …

Your software does not run in a vacuum—the world around it is in constant flux and thus for it to remain effective it must keep pace when the external world forces a change.

If you’re considering building self-order kiosk software in-house, think about what other software platforms it will need to communicate with. Do you rely on outside vendors for your POS software? How about your loyalty platform? Your backend analytics platform? If any of the vendors you work with for these other pieces of software make a change or update to their product, your proprietary software may no longer properly communicate with it. You will need engineers on staff who are kept abreast of these updates and can get to work immediately so that your guests don’t find themselves trying to order from a broken kiosk. Some restaurant brands may decide that is a price they are willing to pay, but anyone making the decision to build in-house should seriously consider this scenario.

But it isn’t only the world around your business that is in flux.

Your needs will change, too.

As Noah writes, the problem with building in-house is “you end up either 1) Writing a specific purpose built solution that will break when you shift your business requirements or 2) Creating a generalized customizable platform that anticipates a broad array of use cases.” Let’s explore this a bit.

In the first case, you have spent money on something that may be obsolete when you inevitably decide at some point down the road that you aren’t indefinitely going to do business exactly as you do today.

In the second case, you need to build a flexible, customizable platform, and that is a very expensive proposition. It is what we are doing at Bite, and it takes a lot of time and effort. It makes sense for us, because we work with a lot of different restaurant brands with different needs and tech stacks that are constantly changing. We benefit from making a tailor-made solution for each of our brands, because there are others out there who can also benefit from that solution. When you are building software in-house to be used by your company alone, this potential upside doesn’t exist.

That’s where we come in.

This is something we at Bite understand very well. As our product has grown in complexity it has naturally grown more costly to maintain—but that is a burden we are eager to bear because this is our core business and our core mission. 

At Bite Kiosk, we live and breathe self-order. We are thinking day in and day out about how we can improve your guests’ self-order experience while in turn providing you a return on your monthly investment in us. We believe that kiosk ordering is the future and it is a future we are ushering in—not just because the numbers make sense but because we believe that our software will elevate hospitality everywhere, empowering guests to order on their own terms and matching the fast pace of their increasingly busy lives.