Data mining in Restaurants

With the food business thriving again in the midst of America’s economic upswing – consistently claiming a whopping 4 percent of GDP — some of the nation’s top eateries are quietly embracing data mining to eke out profit in a tough economy. Timothy is a waiter at the Landmarc restaurant at the Time Warner Center, New York. Timothy has always been a great worker — he clocks in on time and never forgets an order. But his sales of beverages and side dishes were falling short last year. In one month, Timothy has served 426 customers, pulling in $17,991.50 in gross sales with a per-check average of $42.23. That’s $3.84 below the overall per-check average at the Landmarc. It turns out that while Timothy was beating the rest of the waitstaff in add-on sales like bacon or cheese on a burger, he was lagging 2 percent behind everybody else in red wine and liquor sales, and a whopping 14 percent behind his peers in sides like French fries and creamed spinach.

The bottom line was $1,636 of lost sales opportunity in a month — the money Timothy would have made if he’d hit the server average.

We know all this because every item sold at Landmarc — down to the last malbec, martini and red quinoa pilaf — is individually logged and enumerated by a sophisticated software package called Slingshot. The software slices, dices and crunches the data every night, and then serves it to managers with breakfast the next morning.

So when Timothy was up for a performance review last summer, the restaurant’s general manager knew everything about him — information she incorporated into a heart-to-heart talk about improving his average.

Another example would be that of Applebee’s. Data mining has provided Applebee’s with immense insight on how back-of-the-house operations effect front-of-the-house performance.  They were able to track the time it took for certain areas of the of the restaurant’s performance but left many other areas to be explored.  Applebee’s could use the knowledge they have gained to analyze other areas such as support positions, the bar area, and even customer waiting time.

Many restaurants employ busboys, hostesses, drink runners, bar backs, and other support positions for the main staff.  The servers benefit most from this support and sometimes can cater to many more tables than normal because of the support staff.  Could Applebee’s use their data to discover the optimum number of tables for a server at a given amount of time?  Or to find out the number of support staffs needed by the servers?  For example, if there were normally X busboys and Y tables, and then went down X-1 or X-2 busboys,  you would be able to discover how many extra servers are needed. The servers are now going to have to spend more time clearing tables.  Data mining could be used to discover if it makes sense for the bar tender to do his/her own stocking throughout the night or to employ a position known as a bar back.   Also, the information can be use to find out the time taken to make drinks, which can be used to find out the relationship between delayed drinks and delay in service.

The wait time at a restaurant can be a critical factor in deciding if a customer dines with you or goes down the street.  Often times a wait is inevitable on weekends or “hallmark” holidays.  Applebee’s could use data mining to discover how long customers wait before they hand back the beeper?  They could test if offering drink or perhaps a free appetizer while waiting is able to captivate the customers if the wait is more than Z amount of time.  Also, Applebee’s could investigate the results of giving customers smaller menus to flip through appetizers and/or desserts during wait periods.  Will it make the customers more likely to order an appetizer or desserts?  Will customers order desserts after the main course because they have been thinking about that brownie sundae for the last hour?

Data mining has provided Applebee’s with many ideas and strategies.  If they continue using their data mining operation they will possibly be able to gain a competitive advantage and hence increase profitability. 

Sources: http://ismteam10.wordpress.com/2011/09/20/applebees-benefits-from-data-mining/

                      http://www.wired.com/wiredenterprise/2012/05/restaurant-moneyball