There are many predictable factors that can influence changes in transaction mix.  The most notable, and frequent, include; pay-day, the day before or after a holiday, and Fridays.  This is well documented and well understood.  What is less obvious is the impact small shifts in transaction volume can have on the experience of all customers.

If we focus on the shifts in transaction mix observed at bank branches and then apply those learnings to a queuing simulation model, we can quantify the impact on customer wait time.  For this example we looked at an average October day and an October payday to see how transaction mix shifted. 

In all 3 bank branches, “pay-day” showed a significant shift away from standard tree deposits to check cashing (Tree Cash on-Us Check/withdraw, Cash not on-Us Check, Cash non-conforming check).  The smallest shift was a shift of 4% of the transaction volume, and the largest was a shift of more than 6%.  With the non-tree deposit transactions taking 40-90 seconds longer to service, customer wait times were negatively impacted.

To demonstrate the impact that a small shift can have, we ran a service and queuing simulation with the following assumptions:  8 tellers, servicing 500 customers with a target wait time of 3 minutes.  We shifted 6% of the transactions from the standard (planned for) transaction length to a transaction type that required 60 seconds longer to complete.  That shift alone will add more than 2 minutes to the average wait time of every single customer that day.  Because of the compounding effect of the extra service time, the expected wait time increases throughout the day.  Customers early in the day experience wait times within the target, but during the second half of the day, all customers are waiting for well over 7 minutes.  So, because some customers and a lot of non-customers decided to demand more teller minutes, the average customer has a poor wait time experience.

The scheduler or branch manager that anticipates these transaction shifts can quickly take steps to protect the service experience of valuable customers.  There are many tactical possibilities, but the two simplest are to adjust monthly staffing to allow for an additional teller during the days/time periods with the greatest demand for teller minutes, shifting that personnel away from days/times with a low demand for teller minutes.  The other tactic is to add an additional queue for non-customers and/or payroll check cashing on pay-days. 

By adding this queue and dedicating one teller to service that queue, the banks valuable customers get the short wait time they expect.  Non-customers would experience longer waits than they are currently.  The longer waits will either be tolerated as a cost of getting a check cashed or convince them to become a customer so that they get faster service (increased sales).