Design

Great restaurant design has always been part art, part science.

But let’s face it: The science side of that equation has often been a bit shaky and underappreciated. Design, after all, is artistic, intuitive and creative in its endeavor to build brand narratives and create great guest experiences, right? Isn’t science best left to the left-brained IT and finance folks down the hall?

Not anymore.

Data science is increasingly seen as a critical tool for helping brands make smart decisions, not just about site selection, menu and operations, but also about design. That’s in part because technology is making it easier and faster to access more and different types of data. And business leaders simply
recognize that in today’s competitive marketplace, attempting to develop, design and operate without good data analytics supporting decisions at every level is risky.

“We’ve come through a decade of an increasingly competitive restaurant marketplace. There are so many new concepts popping up, trying to be the next hot thing, and every step is being evaluated from every angle to increase odds of success. You can’t just wing it in this environment,” says Rob Depp, vice president, brand strategy and retail environments, at FRCH Design Worldwide. “You may be able to if you’re a single-unit independent, but if you’re attempting to grow a multi-unit brand, drive engagement and attract future franchisees, data is important. It creates a filter for and helps to justify decision making. Design and brand experience aren’t and shouldn’t be completely driven by science, but data can help to give rationalization.”

Data, of course, is a large umbrella, and so-called big data — extremely large sets of data from a variety of sources that can be analyzed to reveal patterns, trends and associations — can seem tough to gather and distill down to what’s important and worth acting on. Few restaurant companies have the resources to handle the increasingly massive amounts of data available, particularly as newer sources — mobile apps with loyalty programs and social media top among them — have turned what was once a relatively simple data stream into a torrent.

Indeed, most companies of any size have long used at least some aspects of big data to drive functions such as site selection, marketing and menu development. And as the National Restaurant Association points out in a guide titled “Big Data and Restaurants: Something to Chew On,” big data isn’t really all that big anymore. Thanks to massive gains in computing power, storage space and new software, restaurants of all sizes can harness their data and uncover useful information from their POS, marketing, accounting, inventory and scheduling systems.

The upshot: Even small brands can benefit from the same type of predictive analytics and business insights that the mega chains like Starbucks, McDonald’s and Subway use. And all brands can gain insights on the growing importance of data collection and analytics by studying what the big guys are doing with it.

During development of the Holler & Dash fast-casual concept, extensive front-end research on Millennial demographics, dining habits and brand preferences drove decisions on everything from the style of buildings selected to menu, technology and interior design. Photos by Mark Steele

Big Data Drives Chains

Big data plays a big role in design-related decision making at Subway. Patrick Rose, senior manager of equipment and decor, notes, “Research is at the heart of our company; it informs our menu, design, technology and more. Our decisions are informed by facts and intelligence with specific insights. We collect global data to understand, for example, how we should design our restaurants to deliver an exceptional customer experience and operate efficiently for our franchisees.”

Subway recently worked with FRCH to develop its new Fresh Forward design prototype. Rose says every aspect of the design, from the colors of the walls to the seating plan, is tied to data insights and analytics. A holistic approach incorporating both quantitative data and qualitative research was applied.

“When developing our research to test the Fresh Forward design, we asked questions such as, ‘Do guests come in more often? Were they likely to stay and eat or did they prefer the on-the-go experience? What did they notice about the design?’ If guests didn’t notice a particular design element or it didn’t improve operations, we took it out. These types of insights informed our design decisions, including how our fresh vegetables are displayed, optimal seating layouts for different locations, and the inclusion of kiosks, which we are testing. We’ll continue to analyze these types of data and trends,” says Rose.

Rose adds that Subway is also looking at new technologies to support its data gathering and analytics efforts. “One of the newest tools we’re beginning to use in our research is eye-tracking technology,” he says. “It uses special glasses to understand the journey guests take in-restaurant and how all touch points interact.”

Next-gen emerging brands, too, are finding new ways to make more and better data part of their development and design efforts.

When Cracker Barrel set out to create new biscuit-focused fast-casual concept Holler & Dash as a stand-alone brand, it strategically targeted Millennials and younger age groups. Now with seven units in the Southeast, the chain conducted heavy front-end research on which to build the brand, says Mike Chissler, chief operating officer. He adds that each of its current units serves as something of an alpha test site for ongoing data collection and brand refinement.

“When we first started, we worked with a couple of different companies to pull a lot of data for us on who this Millennial consumer is, what they value, what their dining habits are and how they look at brands,” Chissler says. “That type of data influenced everything from the types of buildings we look for to the interior environment, culinary approach and service style. It was critical to framing up who this brand is and what the experience is, with design being a critical aspect of experience.”

On an ongoing basis, Chissler’s team now conducts consumer research that solicits feedback on all aspects of the Holler & Dash experience — food and service, of course, but also things like flow, finishes and seating styles. And the company culls data continually from social media, adding verbatim feedback and attitudes to the data pool. Chissler personally taps in daily for quick, how-we-did-today insights and relies on an outside marketing firm to provide monthly in-depth social-media performance analytics, which he cross-references against POS data to help identify trends and patterns.

Among several data-driven design changes made to the Holler & Dash design since its launch in 2016 are menu boards. “We added a digital board with better pictures, and it changes throughout the day,” Chissler says. “We’ll be asking questions and gathering data about how the new design is being received on our next set of surveys and continue to fine-tune as the brand moves forward.”

Another change being made: The company is leapfrogging the kiosk trend, which its data previously suggested was important to incorporate, to focus on mobile-device ordering. “Early on, we looked at how people were ordering and felt based on the research that they wanted the convenience of self-service, so we put kiosks in every unit,” Chissler says. “But further research showed us that when we get busy, people will actually just order online via their phone even when eating inside instead of lining up for the kiosk. Moving forward, we won’t be including kiosks in the front of house for that reason.”

Holler & Dash is also analyzing sales data to drive design changes specifically to accommodate takeout and delivery. “Third-party delivery and mobile orders have blown up in the past two years,” Chissler notes. “That’s something that wasn’t in our original design, but the numbers make it clear that we need to go back and make some changes in the kitchen and at the front counter to optimize that trend. And our new restaurants will be designed in a new way, enabling seamless prep and pickup of food to go. We’ll likely incorporate smaller seating areas, too.”

Another standout emerging chain example: Mediterranean fast-casual CAVA, headquartered in Washington, D.C. Ranked No. 37 on the 2018 edition of Fast Company’s World’s 50 Most Innovative Companies list, CAVA doubled its number of restaurants last year to 45 and increased profitability by 15 percent over 2016, according to the report. A data-driven approach, including hiring an internal team of data scientists a couple of years ago, is credited for helping to put CAVA on the fast track and shining a light on opportunities to enhance productivity, food quality and guest experience.

Among the most notable first steps taken by CAVA’s team was to install a network of sensors inside select units that monitor everything from customer wait times to kitchen operations. Analysis of data provided by the sensors led CAVA to make changes to its queuing areas and menu boards, resulting in greater capacity and faster service. Sensors in seating areas also showed usage and behavior trends that led CAVA to rejigger its prototype for different markets. Specifically, seating space in suburban locations was bumped up by 30 percent to accommodate larger groups and lingerers — both needs highlighted by the data. Josh Patchus, CAVA’s chief data scientist, told Fast Company that suburban stores redesigned accordingly saw revenues increase by 20 percent per square foot.

Sensors have also been used at CAVA to track decibel levels, in some cases leading to remodels of the ordering area to put more distance between the register and the serving line where customers interact with staff and make their choices.

“The average restaurant probably isn’t going to go that far, but if they’re willing to invest in such high-tech data gathering, it’s probably a very smart move because the insights they’re getting are incredibly powerful,” says Stephani Robson, senior lecturer at the Cornell School of Hotel Administration. “Those types of sensor technologies are widely used in the retail industry, sometimes anonymously tracking how people move around in the store. One of the things you agree to when you download a company’s app and click accept the terms — usually without bothering to read them first — is that they get access to your information and can track you when you’re in store.”

The interior design of fast-casual Holler & Dash is a mix of art and algorithm.

Art, Meet Algorithm

While some chains’ uses of sensor and other data tracks directly back to design and operational tweaks intended to boost efficiency and guest satisfaction, design is often still left out of the big-data conversation.

That’s due in part to long-standing functional silos and in part to professional predisposition, Robson notes. “Even a lot of brands that are quite well known and that would seem sure to have basic data available to inform design often don’t,” she says. “And a lot of the ones that do often aren’t good at cross-functional sharing of data. There’s little understanding by the IT team, for instance, of why the design team might find certain data sets useful. There are also a lot of designers who bristle at the idea of data-driven design. They’re creatives and not interested in turning design into something algorithmic. They may rely on qualitative research to support design direction but don’t think so much about the role that analysis of quantitative data can play as well. When you’re talking about the level of investment being made to build out restaurants today, particularly if you’re planning multiple units, you can’t do it based on your gut or on casual observation.”

WD Partners, an integrated branding, design, architecture, engineering, consulting and construction services firm based in Columbus, Ohio, makes tapping both quantitative and qualitative data standard operating procedure for its restaurant projects.

“We take a holistic approach,” explains Robert Seely, CFSP, WD’s director of operations planning and design. “We can really dive into the data and the technical side of things, but we then work with our strategists and our customer experience experts on the design side to make sure we’re not looking at either side in a vacuum. There has to be a balance.”

Most projects at WD kick off with a “rather large information request,” adds Jennifer Baxter, senior productivity engineer at the firm. “Our team, in particular, loves data,” she says. “Part of what we ask for when working with existing brands is quantitative, transaction-level data. When we can get that for a set of stores for a year or more, we can roll it up in a number of different ways that help us understand different aspects of space planning design that might be important. A lot of clients think in terms of annual sales, which is great for the pro forma and understanding what business metrics they need to hit. But when we have corresponding transaction-level POS data, we get insights into things like what peak periods look like, for example, and party size and seating type most used. We can plan literally all of the resources within a specific location to make sure that what is needed from a customer experience, operations and business perspective are designed into that location.”

FRCH’s Depp agrees that providing basic data on things such as party size to the design team is important. When not available, or in cases where the POS system doesn’t capture it, the team conducts observational research in the field to get it.

“Occasionally, a client will come to us with the mindset that they’re just looking for the creative side, the design side,” Depp says. “There can be an educational process for them to understand why we’re asking for more technical, quantitative data for things like seating plans. Sometimes they think they know what they want, but the data suggests a different direction is called for. We’ve seen situations where clients wanted banquettes around the perimeter and four-tops in the center. We’ve had clients who wanted to do all six-tops, when analysis of POS data shows a lot of smaller party sizes coming in. That’s concrete, important information that we can design to.”

When working with clients to create new brand experiences, FRCH follows a three-phase approach to research: inform, inspire and validate. Kelsey Chessey, senior brand strategist, notes that quantitative and, in particular, qualitative data gathered at each phase is critical to shaping successful brand experiences.

“In terms of driving informing design, qualitative data is more important than ever in this marketplace,” Chessey says. “That’s where you get to human insights and figuring out how brands connect with real people on a deeper, longer-term level. Part of the reason that we’ve seen a shift towards brands using this type of data to drive design is that they’re realizing that those human connections create a longer-term competitive advantage. They can help them better understand who they are as a brand, how they can fit into the marketplace and offer something unique. Social media is a particularly rich source for gathering this type of data.”

Dawn Arcieri, senior interior designer at Gensler’s Houston office, agrees that a wide variety of research needs to inform every restaurant project but says it’s often limited-service chains that are most tuned in to the need for it. In those instances, she says, it’s usually all about creating a great guest experience while at the same time getting people through the line faster.

“In that regard, we look for a lot of different data points,” Arcieri says. “For instance, how long are the wait times? How are people moving through the space? Are people coming in in groups? How large and where is the queue positioned, and is it cumbersome for diners? If we’re able to get solid, data-based answers to those questions, whether via sensors or video or other observational research, we can tweak the design to make it better. Most operators, however, haven’t yet started to drill down into that type of data.”

It’s clearly not just limited-service restaurants that could benefit from a more data-driven approach. Information that’s readily available to all brands often just isn’t tapped or analyzed for applications beyond operations and marketing. That’s a missed opportunity, Arcieri says. Online reservation and seating systems, for instance, show party size, wait times, requests for seating in particular areas or for particular types of seating, all of which can inform a host of design decisions when that type of data is tracked over time.

Gensler recently designed a new prototype for Burger Boys, a legacy brand in San Antonio whose new owners plan multi-unit expansion. While the operator wasn’t able to provide much in the way of quantitative data beyond sales on which to base design decisions, Arcieri’s team undertook extensive on-site and customer research to determine how best to give the iconic 1960s, 800-square-foot, 8-stool burger joint a fresh new look and add operational efficiencies needed to handle the 1,000-plus-burger-a-day volume.

“Sales data showed they do a ton of drive-thru, and simple observation showed that people had to stand in lines between the eight stools at the counter to place their orders. There was no place to wait,” Arcieri says. “We ended up doubling the unit size. We added space for interior seating and created a queue line that follows the outer wall of the building and comes around to the counter. We also increased the kitchen size and expanded the drive-thru area. The redesign opened about 30 days ago, so we’re now collecting a lot of quantitative and qualitative data on how it’s performing, which will be used to inform the next unit’s design.”

The big trick with data, no matter where it comes from, is ferreting out the nuggets that matter most for driving specific design decisions. And with ever-increasing amounts of data available, it can be tough to avoid what WD’s Seely and Baxter call “analysis paralysis” and/or the temptation to over-operationalize, ultimately losing sight of the real consumer and the experience you’re trying to create.

“It’s pretty easy in this data-centric environment to fall into over-analyzing and relying on data when sometimes the best insights can be gained by simply engaging with and observing customers,” Baxter notes. “You need a balance of ‘here’s what the data tells us’ and ‘here’s what consumers say, or what we know about them from qualitative research.’ It all has to work in tandem to create really good designs and really good experiences.”