Reducing wait time: Can data analysis get your latte to you faster?

Barb ClearyCan data analysis get your latte to you faster? Restaurants and coffee shops are discovering the power of process analysis in improving service and delighting customers.

The last time you realized that the line you chose to queue up in at the bank or coffee shop had become the slowest one, while other longer lines moved quickly, you perhaps wondered about improvement of processes in service environments. It may seem easy to envision data analysis utilized in manufacturing a what-zit, where defects may be easily spotted. What constitutes a “defect” in a wait line for a bank or for a meal at a restaurant?

Perhaps you have also noticed that at your favorite coffee shop, drive-up customers seem to move through more quickly than walk-ins. While you stand at the counter contemplating the amount of fuel consumed when a dozen cars wait in line outside, the twelfth car drives up just as your order is finally taken. Data collection and analysis could surely improve this system, which leaves you unhappy with the coffee shop and determined next time to drive up rather than take steps to go inside—if you visit this coffee shop again.

Restaurants and coffee shops have indeed pursued data analysis in order to improve processes. Nearly a dozen years ago, for example, Sudie’s Fish House in Houston pioneered the use of CHARTrunner to collect and analyze data related to time from plating food to delivery at tables. Inspired by the need for hot food for customers, Sudie’s owner, Paul Bailey, charted the process and reduced wait time and improved customer satisfaction.

Sudie’s developed a clear process for measuring wait time. In the kitchen, completed orders would be tagged electronically. If they were not served within 8 minutes, a blue light would come on. For 15 minutes, a red light would illuminate. Waiters and kitchen staff became acutely aware of their orders, and data was collected over time and analyzed. Delighted customers commented on the improvement, and the restaurant benefited from faster “table turns,” since things were moving more quickly.

Clearly, the analysis of data related to wait time involves the application of other process tools. What constitutes an excessive wait, for example? An operational definition, formulated by evaluating the process, will establish limits and identify the measurement process. Sudie’s used 8 minutes for the first alert, recorded electronically and alarmed with a blue light. Prior to developing this definition, it would be necessary to examine the system as it currently operates. This means collecting data to reflect current wait times.

In a coffee shop, this data collection might look something like this:

“Wait time,” however, must be carefully defined prior to collecting data. Does it include the time that a customer waits in line to place an order? The time it takes to prepare the coffee drink? An operational definition of “wait time” might be “In minutes, the amount of time elapsed from a customer’s approach to the counter to the moment when the coffee is delivered, as measured by timer located at cash register.” All those who collect and analyze data must be clearly aware of the ways in which wait time is actually measured, and must record it consistently. If one server begins the count the moment a customer enters the store, and another starts when the customer gets to the counter, the data will be inconsistent and inconclusive.

In addition to collecting data on wait time, restaurants can develop charts of typical customer complaints, and using the Plan-Do-Study-Act cycle, respond to these in meaningful ways. What might be the greatest customer complaints? These Pareto charts indicate not only the number of complaints in each category, but the cost that accrues to each of the complaints. (Costs must be determined by calculating the time involved, the likelihood of a complaint determining customers’ future visits, etc.).

Note: “Total cost” was calculated using a formula that included resources (time, financial investment, training, etc.) that are demanded in order to improve the situation (not necessarily “solve” it entirely) and satisfy customers so they will return to the restaurant.

While it may not be easy to classify something as “defective” in providing service to customers, it is clear that negative experiences will affect the organization’s profitability in the same way that poor quality affects manufacturing outcomes. A variety of metrics can be gathered to indicate good customer service—the number of times customers return to the restaurant, for example, or the comments that they may make to their servers. These may seem to be too subjective for measurement, but with the right approach to operational definitions, data collection techniques, and clear analysis, service organizations, from restaurants to hospitals and banks can use data analysis to evaluate the customer experience.

And regardless of the industry, the response of the customer is what will drive profitability.

Barb Cleary

Barb Cleary