Queue-busting at banks reducing

The challenge

No one likes to queue. A research agency had collected extensive data for their client, a national bank, on precisely this issue. They had data on the length of time customers spent waiting, the number of staff at the tills and other relevant information. Data was available for many different branches of the bank for countless hours.

The agency outsourced the statistical analysis of the data to us.

Objectives

To identify the factors influencing queuing time so that queues could be kept to a minimum by allocating an appropriate number of staff to serve on the till.

Our approach

We created a statistical model using the data provided by the research agency. This included tens of thousands of observations, showing queuing time and the circumstances at the time.

Outcomes

Our analysis explained a large part of the variation in queuing time. Local factors such as pay-days, market days, levels of income and number of staff servicing the tills had a huge influence. More general factors such as day of week, time of day, seasonality and the state of the economy also affected queuing time.

We made recommendations as to the best number of till staff at any particular time. These varied depending on
different assumptions as to what was an “acceptable” time for a customer to queue.