Predictions are always suspect, but some trends are undeniable. Automation has permeated our culture since the early days of the industrial revolution. The formula is simple - if an industry relies on processes that can be systematized, repeated, and scaled profitably, it will eventually be "disrupted". This is sometimes called "progress", a Darwinian drive to remove inefficiency in pursuit of profit. Do more work. Do it faster. Do it cheaper. Rinse and repeat. Computers have accelerated this, and they consistently get faster and better at it. But one area of everyday life has stubbornly remained immune from automation: the full-service restaurant.
Computers are great at repeatable processes given predictable inputs. We're currently experiencing a revolution in AI. Systems that, given unpredictable input data, can produce surprisingly predictable results. Machine learning is exploiting the computer's ability to test models very quickly, iterating over input data millions of times, creating and scoring models that produce a desired goal until the best results are achieved. As a result, we're seeing systems that can easily understand and generate speech, identify faces in video and images, predict buying patterns, identify emotions in speech, and more. But what machine learning needs to work effectively, is data, and lots of it. Full-service restaurants have traditionally been data-poor environments.
This resistance is primarily in the "front of the house." The back of the house is ripe for efficiency tinkering, since it's hidden from the public eye. But the front of the house, the final product/market interface, is constrained by tradition. People in western countries have a lifetime's worth of expectations about what happens when you set foot inside a sit-down restaurant. It's a ritual people have grown up with. If it's a fancy place, you probably made a reservation, so you need to let them know who you are when they ask. You are greeted by a human host, and guided to your table. You are presented with a menu, and place your drink order. Your drinks arrive, and you place your food order. Your food arrives, and you (hopefully) enjoy it. Your server is generally expected to stop by and "check in" on your meal after you have had a chance to sample it. You are presented with a check, which is processed by a human. If your experience is satisfactory, you tip your server. This ritual has a lot of variation across establishment types and markets but the general framework is recognizable across most types of full-service restaurants.
It's a "target-rich" environment for automation. Repeatable steps that generally follow the same pattern from restaurant to restaurant. Each step can be put through the "progress" mill we described above. Let's envision a fully automated Restaurant of the Future. You make a reservation in the restaurant's app, dinner for two on Saturday night with a view of the water. Upon entrance, your phone pings, alerting you that your table will be ready shortly. After a short wait, it pings again, this time with a picture of your table, informing you that your table is ready, and that you can seat yourself. A map on the phone in the restaurant app shows you where your table is, so you head over, guided by a map of all the tables. Once seated, the table itself has a discrete "smart speaker" which you interact with, placing your drink order. The speaker's persona is friendly and personable, indistinguishable from a real person. The persona knows your preferences, allergies, and any food restrictions you generally make public, and discreetly recommends specific selections it predicts you might enjoy. After discussing the day’s specials, you review the standard menu on your phone, and place your order. Food delivery at this establishment is by robot waiter, as human service is only available at very fancy places these days. After an enjoyable meal, complete with dessert and after-dinner coffee, you leave the restaurant, your bill being served via your app. A few days later there's even a follow up call from your "waiter", asking how your experience was, and offering a discount on your next visit.
Dissecting this science-fictional scenario is easy. Can it be done today? Mostly, as many of these systems currently exist. Geofenced apps can trigger confirmation of arrival for a given reservation, easing users into an in-restaurant journey. Interior mapping solutions exist, for guiding users to their table. Data mining powers predictions based on past behavior. Smart speaker agents like Alexa, Google Home, and Siri already enable basic imperative communication, and in the near future will become indistinguishable from human speakers. Paying for your meal with your phone is already old hat. The only real "stretch" in this scene that doesn't exist in nascent form today is a food service delivery robot flexible enough to handle the myriad items that human waitstaff can intuitively do. Replacing staff with machines is the inevitable collateral damage of automation, and attractive to owners because machines don't break down, have fixed costs, are generally cheaper the more of them you buy, and never show up late to work. Everything in the above scenario seems inevitable, and just a matter of time. Someday, only the most exclusive and cheapest restaurants will employ human waiters.
We're already seeing it. Most "quick service" restaurant chains either have already deployed or are investing in mobile ordering. Many "family restaurant" chains already have multi-functional tablet kiosks where you can order and pay for your meal, complete with games to keep the kids busy, surveys about service and food quality, and on-ramps to join loyalty programs. Larger restaurant chains are increasingly investing in consumer data platforms to help them identify their best customers, create repeatable engagement patterns, and reach loyal customers with offers to keep them coming back. Leveraging technology in the dining experience enriches the data space available to such systems, allowing machine learning systems to create predictive analytics, insights, and customer scoring across many data points. The more you know about your customer, the more prepared you can be to make informed business decisions, market effectively, and keep them coming back.
At SessionM, this is our sweet spot. Taking disparate data sources into our platform, building repeatable user flows that reward engagement, and producing actionable user scores is central to our vision of the future of loyalty. We have no idea if the futuristic restaurants staffed by gleaming robot waiters are just around the corner, or if they would be accepted by diners. We are, however, certain that as more technology is deployed across the dining experience, there will be a need for systems like ours to make sense of, and more importantly, effective use of, the resulting avalanche of data. Contact us to find out how.