Home / Banking Strategies / Improving Teller Performance with Peer Data

Improving Teller Performance with Peer Data


For many banks, improper scheduling on the teller line is one of the biggest silent profit killers in their retail branch network. Because it is difficult to precisely identify exactly what each teller is doing and how this correlates with account holder traffic flow, many financial institutions (FIs) persist in scheduling based on what is available to them – best guesses.

A solution is available, however. Gaining access to the transaction data of peer institutions enables FIs to benchmark teller performance effectively, both initially and during the course of efforts to improve teller efficiency and performance.

Data Conundrum

Experience shows most FIs handle teller staffing using one of four models:

  • Asset Size: determining staff size based on the asset level of each branch;
  • Budget: budgeting staff expense to be in line with previous years (plus 3% for inflation);
  • Capacity: staffing every available teller window in the branch network from open to close; and
  • Demand: scheduling both full-time and part-time tellers to be in line with accurately forecasted transaction volumes and traffic flow patterns per branch.

The problem with the first three approaches is that they are not closely tied to actual teller performance—or even business forecasts. As a result, there is little way for banks to determine if the schedules are optimal, or if tellers are being productive during their schedule periods, whether it’s servicing account holders or carrying out other sales and customer support duties. This use of disconnected scheduling methods is precisely the reason many banks are unaware of how much poor teller productivity may be costing them in either excess staffing dollars or not properly redirecting tellers during the non-customer facing times towards more precisely planned sales and customer support tasks.

In many cases, these “best guess” scheduling approaches leave managers blind to whether or not their tellers are producing at desired levels throughout most of their shifts. With the same process in place going forward, they may be perpetuating the subsequent results from their improper scheduling conditions.

Ironically, the fourth approach – the demand model – is by far the most effective, yet it is also the most underutilized. Banks may not understand how it can benefit them. Using peer productivity performance data eliminates this conundrum by giving banks hard data, based on real-world FI performance, against which they can measure their own branch network. It validates the premise that increased teller productivity is achievable and helps managers identify the metrics that will be meaningful in those efforts. It also helps them set realistic productivity goals, which is important to obtaining employees’ buy-in for improvement efforts and encouraging them to strive for higher performance levels.

The first step is for a bank to identify a data source that contains some measurement of teller productivity, such as transactions per hour (TPH) worked, outbound sales calls, or labor cost per-transaction (LCPT). Financial Management Solutions, Inc. (FMSI) defines TPH by how many unique monetary transactions a teller performs each hour, measured in 15-minute increments; if no transaction takes place in a 15-minute time increment then this increment is not counted. LCPT is obtained by dividing the employee’s payroll rate, or salary and benefits, by TPH.

The bank then obtains both internal and external branch data measuring the same indicators. Having parallel metrics from the different data sets is very important to this effort. With this information in hand, banks can begin making comparisons. The metrics may change, depending on what mechanism you use to gather and calculate the data. The important point is to compare like datasets so you can see where you are and how much improvement you might be able to accomplish.

In our experience, the differential in average cost per-transaction between the top 10 and the top 50% of banks is significant, approximately 17%. Typically, banks that have not initiated any scheduling or performance optimization strategies, based on forecasting account holder branch visits, have much lower teller productivity rates than those that do. The labor savings from even an incremental improvement can save a mid-sized institution tens of thousands of dollars per month.

At this point you may be saying, “Sounds great, but where do I get this productivity performance benchmark data?” Although some survey firms produce limited pools of peer transaction performance data, the easiest option is to work with a resource that harvests transaction data directly from multiple FIs.

If you have a strong relationship with another bank or other FI, ask if they have a solution (developed in-house or through a third-party tool) for harvesting core processor teller transaction data. If so, they may be willing to share data (scrubbed for anonymity) about transaction productivity rates at their branches and even to make referrals to help you collect yours. If not, a few companies have developed tools and software to help banks harvest this data within their own branches. A number of vendors also offer historical peer data, collected for their clients, that they will share anonymously for benchmarking purposes during free consultations or evaluations.

To harvest your own data, we recommend a dedicated core-processor harvesting solution, developed in-house or via a third-party. Some banks also obtain data by observation. This method is imperfect and practical only for limited data collection; it works best through the use of video cameras in the teller area.

A third possibility is to benchmark the tally of transactions per shift in your different branches. This can give you limited insight into teller productivity over prolonged periods. In any of these cases, as always, make sure your peer data or the different branch data mirrors the same metrics for comparison purposes.

Mr. Scott is president and CEO of Alpharetta, Ga.-based Financial Management Solutions, Inc. (FMSI), which provides financial institutions with business intelligence and performance management systems for efficient staffing. He can be reached at [email protected].