Not so many years ago, big data was defined as anything too big to fit in an Excel spreadsheet. For restaurants, it generally meant financial information piped in daily from cash registers. These days, point of sale no longer is enough. When Domino’s alone collects intelligence from 85,000 sources a day, small and midsize operators, such as Chronic Taco, have to boost their data programs by pulling in new sources of information in order to compete.
“We find a new use for [our] system every day,” says James Boswell, director of financial planning and analysis at 700-unit Marco’s Pizza, headquartered in Toledo, Ohio. “Our goal as a business analytics team is to provide an answer to whatever questions the business asks.” And that means using data to inform all aspects of the operation, beyond finance to menus, marketing, store design and more.
So what are the must-haves for upping your data game today? It’s not just a matter of more data. The trick is finding the right data and interpreting it in the right way to make informed decisions that will improve business. We asked information gurus from a variety of restaurants to put together a picture of Big Data 2.0. For any size operation, here are their tips for getting more out of the numbers, as they grow bigger and bigger.
1. Involve everyone
When Danny Walsh joined Jacksonville, Fla.-based Firehouse Subs as director of reporting and analytics, one of his first moves was to convene a data summit. He asked his CEO and other officers what questions they wanted big data to answer, and what measurements might be key to those answers.
Once the new system was ready to roll, he trained regional managers and deputized them to train subordinates. Anyone with questions could browse a video library.
Thanks to that groundwork, its system, called Station Pulse, gets used at every level of the company. “I’m amazed how much this data has become a part of daily routines,” Walsh says. “It’s used in all types of meetings, from the C-suite all the way down to the franchisee or general manager.” And they’re discussing not just significant numbers, but how to improve them. For example, using data helped the chain shift efforts to increase contributions to its Public Safety Foundation charity an average of 22% per store over the previous year.
The ultimate goal is to put data at the back of everyone’s mind, says Chief Data Scientist Josh Patchus at Cava Grill, a 19-store fast-casual Greek concept in Washington, D.C. “You have to breed a data culture,” he says. “You have to make sure people understand and appreciate data and use data in everyday decisions. Make sure that every new hire has a data-driven mind.”
2. Speak one language
“The biggest part of putting our system together was data wrangling,” says Boswell. “It’s a simpler task to put business tools on top of a database that’s already laid out correctly.”
Before Marco’s Pizza’s sales reports, demographics and Facebook comments get fed into his data warehouse, they’re translated into a common format. Boswell worked with two different point-of-sale vendors to make their records compatible. Such “normalization,” he says, allows non-experts to query the data and get answers. As a bonus, the architecture is malleable enough to add in new datasets such as on-the-fly maps and weather, allowing for future innovations to inform the business.
Having diverse information all speak the same language lets Boswell tackle questions like how to increase online orders, which carry higher average checks than those in stores. He is able to meld sales information with demographics, online comments and web interests to learn what motivates online diners.
He also can explore more unusual queries, like whether pizza orders go up when there’s a political debate on TV. It turned out that last fall’s first two primary debates saw spikes of up to 40%. That prompted Marco’s to add workers and do social media promotions around later debates. “That’s all possible because we have our data in order,” says Boswell.
3. Organize feedback
Big data is able to look beyond a list of Yelp reviews or Facebook comments. Today’s programs can aggregate hundreds of customer comments in a week to give a real view of what customers think—and soon after they think it. “If you can see the big picture, it tells you a story you can act on,” says Lesley Truett, vice president of marketing for Rosa Mexicano, with 18 stores based in New York City. “Your guests are giving you the answer.”
Instead of spending time searching the many different sites, she has a third party gather reviews daily across the web and sort them by topics such as food quality, service and presentation. Once the data pinpoints an issue, she can work out solutions. For example, when guests on many different sites complained that enchilada portions were too small, she changed the plate size and added a third enchilada—while keeping the same amount of filling.
Understanding the results doesn’t have to be high-tech. At Chamberlain’s, a concept comprising a steakhouse and a seafood spot in Dallas, owner Jeffrey Barker sits down every morning and reads a compiled digest of online comments, culled from a range of digital sources. When he comes across a negative one, he’ll pull up last night’s sales records to learn more about the visit before he takes action. “I can see what they ate, what they drank, when they got there and if they exaggerate,” says Barker. “The computer doesn’t lie. Then I can call the guest.”
4. Appeal to the eye
When a store manager at Dickey’s Barbecue Pit logs into its data system, dubbed Smoke Stack, they’re presented with a dashboard. In a dozen graphs and tables, it offers a snapshot of the business for the day, the month and the year: what’s selling, comp sales, labor costs and more. If the manager spots lackluster numbers on catering or complaints, he can dive deeper and create custom graphs from the data in the system. “The secret to making data actionable for anybody is to pull them into it visually,” says Laura Rea Dickey, chief information officer for the 562-unit Dallas chain.
All figures on the screen get updated every 15 minutes. While tracking a new taco, she says, “I can look at sales in near real time across the system and overlay them with other data points, like how it’s trending on social media or how it’s doing in markets with media campaigns compared to markets without.”
At Firehouse Subs, the dashboard paints an old-school kind of picture: a report card. “We needed a one-stop shop for telling the story of what’s going on in a particular restaurant,” says Walsh. “What better way to visualize than with a scorecard?”
Its Station Pulse system rates each store on nine key performance indicators. Distilling the data is crucial, Walsh says. “You can get into the trap of putting too much clutter out there. We want to focus on data that can provide actionable insights.”
Of those measures, half of a store’s grade comes from financial numbers, including catering dollars and comp sales. The other half comes from computer analysis of guest surveys, coded for levels of favorability. When the report card calls attention to an indicator, Walsh says, operators tend to focus on and improve it. In the first six months of using the system, ratings for both table touches and overall guest satisfaction rose three percentage points.
5. Big data at little cost
While many of the different APIs on the market can seem daunting, and the idea of building a data system from scratch can sound pricey, getting a data program in place doesn’t have to mean nightmares of dollars signs. At Chamberlain’s, the net cost of its data program is zero. When a new credit card processor wanted Barker’s business, it threw in an array of data analytics—and guaranteed the same rates as his old vendor. “I’m getting a little bit of a gift,” he says.
He’s not the only operator who spends small money for big data. In Denver, two ViewHouse sports bars spend $200 a month on complimentary Wi-Fi for customers who provide their names, emails and birthdays. Depending on how often they visit, says Marketing Director Jennifer Ruppert, “We can retarget them with an email if they haven’t returned in a month or six months. It’s an easy way to turn customers into repeat visitors.”
Ruppert also exploits the analytics that come with Facebook business accounts. She harvests ages, genders and locations of people who have “liked” the bars’ pages. Then she’ll post an event and spend $50 to promote it—not just to those fans, but to other Facebook users with comparable profiles. One post can reach up to 20,000 people, including travelers from other cities.