Metrics in the Mix at Figi’s

Problem: Call volume that spikes from 60 calls per day during non-peak season to 60,000 calls per day during the holiday rush.
Solution: Use metrics and flexible-scheduling techniques to help achieve agent efficiency without overstaffing.
Results: 97-percent call center service levels and 90-percent agent-utilization rates maintained throughout most of the rush.

As it is for many catalogers, the December holiday season is a booming time at Figi’s, the Marshfield, WI-based gift-giving speciality food mailer. Call volume builds from 60 calls per day during the non-peak season, to a whopping 60,000 calls per day during the holiday rush.

To handle the volume, Catherine Paul, call center manager, expands the year-round staff of 100 associates to 1,200 in two call centers. Following are some of the ways the customer service staff is ramped up to handle the peak.

Flexible Scheduling

The first objective is to determine how many seasonal agents to hire and by when they need to be trained. For example, when planning for the 2002 holiday season, Paul calculated that in the peak week in December, she would need 37,000 staff hours of call-center coverage. Each agent typically worked an average of 31 hours per week.

So 37,000 divided by 31 equals 1,194 agents needed on the phones that week. She also did this calculation for anticipated peak weeks in September (111 phone agents needed), October (247 agents) and November (395 agents).

Working backwards from these numbers, Paul and managers from Figi’s customer service, training and operations departments devised an efficient training and subsequent work schedule for incoming seasonal agents. They worked from what metrics they knew, namely how many staffers Paul needed on the phones and by when.

The call center also uses flexible-scheduling techniques. For instance, when seasonal workers are hired, they’re asked for their preferred and available hours. Preferred hours are those the employee wants to work, and available hours are those the agent could work if necessary.

This information then is loaded into Figi’s workforce-management software, which holds these two types of data separately. Paul says she schedules agents first from their preferred hours and then from their available hours.

Other flexible-scheduling tactics include what she calls CYWFM, or “Can You Work For Me?” forms. Agents can post their assigned hours if unable to work them.

Lists of agents who are on-call if additional projects or heavy call volume comes in are created across all customer-service areas. Seasonal staffers also are trained to handle customer correspondence via mail and e-mail, and to do returns processing.

To help discern Figi’s aggregate call-center staffing needs, Paul uses numerous metrics such as catalog drop dates, usual call-to-order ratios, average call length by type, and anticipated call volumes per day and per week based on historical data and marketing forecasts. Additionally, she tracks and monitors actual call volume and makes staffing adjustments as necessary.

All of this is accomplished with the objective of maintaining Figi’s corporate goals of a 97-

percent service level (3-percent call abandonment) and 90-percent agent utilization (on the phone or in after-call processing for 54 minutes per hour) during peak season.

“Everything we do, every decision made, every forecast — it’s all based on reporting and measuring,” says Paul.