BB&T Optimizes Teller Scheduling with GMT Workforce Tool

Finding the right balance when staffing branches can be a tricky, sometimes subjective task. When Winston-Salem, N.C.-based BB&T ($154.2 billion in assets) saw its existing system for teller scheduling was not as optimized as it could be, it called upon the expertise of GMT Corp. (Norcross, Ga.) to remedy the situation.

Like every bank, BB&T has focused on gaining efficiencies. According to John Charles Thompson, EVP, manager of branch operations for the bank, this need was even more apparent at BB&T given the number of acquisitions it had performed over the past two decades—60 community banks, more than 85 insurance agencies and 35 non-bank financial services companies. This resulted in there being little standardization from branch to branch and region to region.

"We saw we were overstaffed at the teller line," Thompson says. "We had a variety of branch types with unique challenges around staffing."

Since there was no centralized teller system at BB&T, the bank relied on data collection at the branch level to determine the teller schedule. Reports were produced at the end of each month tallying the average transactions per teller. This was compared at each branch and by each city/region, but little was achieved in the way of efficiency with such data.

"There was no degree of scheduling or forecasting assistance," he says. "It was just one dimensional, historical data. There were no sophisticated modeling tools to staff against forecasted needs, so we needed to do something."

Thompson says when he entered his current role in 2005, his main focus was looking at the bank's internal systems for more efficiencies and he embarked on this current campaign in 2007.

Thompson's plan for seeking teller scheduling nirvana involved two very important points: gaining the efficiencies the bank felt were available and ensuring BB&T still maintained the highest quality service possible using the right number of tellers at the right time.

"This required a forecasting tool," he explains. "Our operating model is a high quality service model and we didn't want to damage our service. We usually get good scores from third party marketing companies on customer satisfaction at the teller level. So we wanted a solution that gave us ease of integration into our existing [ARGO] teller system, and ease of operation so it was not just efficient, but simple to use. We didn't want the branch people spending a tremendous amount of time learning this solution."

As a result, BB&T issued an RFP at the end of 2006 and made its selection of GMT in the first quarter of 2007. Implementation began that summer.

BB&T examined three or four vendors, including GMT, ARGO and Exametrics, but ultimately decided on GMT due to its ease of use and the manner in which it integrated with the bank's systems, according to Thompson. BB&T uses an IBM mainframe along with HP servers running SQL and DB2 for its databases. Storage is handled by Hitachi systems. With the new implementation, scheduling is now done at the branch level with the GMT Planet workforce optimization solution operating as a Web-based tool running over the branch intranet.

The implementation took place over a three-month period. In addition to receiving help from GMT, a third party consultant, CAST Management Consultants, was also on hand. CAST's role was to perform time studies in a sampling of BB&T's branches. The data produced by these studies was used to feed the GMT forecasting tool. CAST looked at all the activity in 20 branches of various sizes and in different locations for a month in August 2007. The consultants gave tellers a "placemat" of the bank's services and they would scan a given service with a barcode scanner each time they performed that activity for a customer. They would scan at the beginning of the transaction and at the end. Unique time standards were developed for each transaction activity. All this data was then transferred to the GMT solution. As a result, BB&T has had 90 percent forecast accuracy with the workforce optimization tool, says Thompson.

In November, the system was turned on and tested for several weeks. In February and March of 2008, BB&T rolled the tool out to all 1,500 branches.

"It was a very quick rollout," Thompson recalls. "By the end of 2008, we achieved our stated projected goals: 5 percent reduction in teller line resources and we had tellers scheduled at the right time with no degradation in client service quality and no layoffs."

As with any new technology, there was a learning curve. However, once BB&T rewrote the training materials, things went more smoothly. "We didn't test or pilot it for very long—about 30 days," he explains. "BB&T has its own training division called BB&T University. So we wrote our own materials and repiloted the software after another 30 days. But we were able to roll the solution out to our 11 states, 33 regions and 1,500 branches in just four months. We are live and complete. We didn't hit the ground running at 100 percent efficiency, but it's getting better each month."

One of the measures included in GMT Planet is "workload fit," meaning pairing the right personnel scheduled at the right time of day. Thompson says BB&T is able to accomplish this 70 percent of the time now, but the goal is 80 percent. Managers are also able to post schedules three weeks in advance, a vast improvement from the previous method. There is now more standardization of hiring, scheduling and management of teller personnel throughout BB&T's banking network. Additionally, BB&T saw a reduction of teller overtime compensation by 50 percent. The bank also expects to save 15 percent in labor costs over the next two years.

There are other benefits to the workforce tool as well, such as lobby management—the branches are able to infer how busy they are at particular times of day.

"We feel our employees develop relationships with clients," Thompson says. "It is critical that we ensure this. [With the tool], we're enjoying less turnover than in previous years at the teller line. This may be due to a number of factors, but I think the more someone knows their schedule in advance, the happier they are."