Wikia

Call Centers

Forecasting

Talk1
7pages on
this wiki

The Importance of Forecasts How often have you seen or heard about call center management struggling with the senior executives or their own staff about the resources needed by the center to do the work. Missed service levels, tired or worn out staff, blocked calls, long hold times are but a few of the issues facing call center management everyday. Is there a better way?

Fortunately yes! One way is to have a forecast of work volume and staff for at least six months by week and eighteen months by month. This type of forecast is done with a few hours work and a spreadsheet program. Like most systems in call centers today it requires input from a number of sources; sales, marketing, operations and call or work history from the center itself.

To be sure on the terms, this is not a staffing scheduling forecast; one that is generally between one to five weeks out. That type of short term staffing forecast is fine and is needed to plan the shifts and work arounds. Staff scheduling is tactical and day to day. It does not provide the kind of insight needed or desired by senior officers of any organization.

There are many applications and tools out there now that provide a very decent calculation of the number of staff needed to provide a given level of service. These tools and applications support the creation and maintenance of staff scheduling. For many centers this is the point at which they stop.

For a few, it is the starting point. From here the few gather other key information. Such as: volume by period for the last 18 months by transaction type, this shows any seasonal variations and/or trends that need assessing; rate of absents, sickness etc., this allows a judgment of extra or additional capacity from existing staff that is likely to be needed; vacations, how many man-days and when, (service still needs providing, but staff seem to like vacations, go figure); special events or marketing initiatives likely and the range of effects on the work load, while not exact, something is better than nothing.

Marketing and sales will be pleased that you cared enough to ask. Be aware that many “forecasts” from marketing and sales are more likely goals and objectives for the campaign. Be sure to check past campaigns for projections and actual results. You may need to develop a ‘discount rate’ to view actual results versus planned goals or targets. This gives a good indication of the range that is possible not just what someone’s budget goal for the year is projected at. At the same time get sales to provide the run rate of the last 12 to 18 months and the plan for the upcoming period. Once the sales and marketing historical data is provided it should be checked and matched against the historical call rate for the center to establish whether or not there are any relationships. A common relationship is a lag between a sale and the service call. Another common relationship is the delay between placing an order for a product or service and its delivery. The greater the time between these events the greater the number of calls a center generally receives.

Any training and special projects need identifying. Training for both new staff (induction) and on-going refresher or upgrading for existing staff. Keep in mind the churn or turnover rate for the center. High growth or high turnover either internal, churn or people leaving cause productivity decreases. A ‘Special projects’ allowance is always good to have in reserve as well. These projects provide staff with an opportunity to demonstrate skills and knowledge outside their usual arena. This can be a key element in staff development and they are a very good morale booster for everyone.

From these basic measures of work and capacity, a staff full time equivalent (FTE) requirement can be determined. Depending upon the center policy toward full, part-time, contract staff mix will dictate the number of total staff and the number of seats needed for the call or work demands by period.

The demand is the call or work load, the number of calls likely per period. It is essential that for budget purposes an 18-month view by month with assumptions be the start of the forecasting process. Once completed it is easy to keep up-to-date on a monthly basis with weekly reports of actual versus forecast. Eighteen months is roughly one business cycle for most companies or organizations. It is not too far off in the future to fall suspect to wishful thinking or idle speculations. It is something that everyone can see and understand is in the near term. It is also far enough out that proposals regarding capital expenses can be developed and presented for approval with a fair chance of decisions before the projections become invalid. This improves the business planning process. This view also means that there are no surprises about changes to volumes at budget time.

It is a good idea to review the data, assumptions and the methodology approach with the senior officers and those in the operations. This gets them comfortable with this type of approach; builds the communication and understanding of why the center operates the way it does. With this in hand any center executive can discuss with everyone, staff, operations, HR, senior officers any aspect of the center based on the work the center is asked to handle. Everyone then is working from the same information and knowledge.

Once agreement is reached on the forecast it is then deconstructed to the operational forecasts for block scheduling and staffing requirements. First calculate the staff hours required to handle the projected work for an operational 52 week forecast. Identify any periods of very high volume or very low volume to block out best times for vacations, special projects, new staff likely hiring, training periods and black-out weeks (i.e. no vacations, special events or training, anything that might reduce the capacity of the center to handle the work). Add the allowances for total vacations earned or expected, training, special projects etc. Use the factors for churn and turnover to identify the number of new staff and the best periods to recruit and training. Add all these together and make an allowance for sick and absences, reviewing past data to see if there is any seasonality that should be taken into account.

The result is a 52 week forecast of staff hours needed to successfully run the center meeting all service target requirements. This is then broken into a model of full and or part-time headcount requirements. Each center has it own approach and ratios of full and part time mix. Generally centers try to find the mix that maximizes their flexibility to grow or shrink the hours to meet the work and demand changes.

For instance if a center requires 30 FTE (full time equivalents) it might have 18 full time staff, 12 part-time, 8 on contract or temporary. This totals a headcount of 38 but with the same base hours as 30 full time staff. This practice provides flex for the center manager to change capacity and optimize staff mix as demand ebbs and flows. At any time the hours available could vary by 33%.

This ability to adjust and react is critical. The operational managers of any center are always trying to produce the required service at the least cost. It is easy to reduce the costs in the short term by reducing staff headcount to the bare minimum required to meet the service now. This strategy is short-sighted and counterproductive. Reduced staff burnout faster, turnover increases, service levels degrade and AHT increases due to poor service. A strategy of adequate staff with flexibility to adjust to meet both work and service demands keeps both the service quality and costs in-line.

With the staffing requirements completed scheduling is the next step. Schedules should generally be done between 2 to 6 weeks in advance. Adjustments below this level can easily be handled with overtime, extended hours, reduced work etc. between the managers, supervisors and agents. Center managers keep an eye on the total hours, calls and the budget for the period. If they are in balance all is fine. Adjustments or trends are noted and tracked for inclusion in the next revision of the forecasts and operational schedules.

Forecasts are part science and math, part experience and judgment. Use, understanding and experience drive revisions. These require posting and publishing to everyone concerned once approved and accepted by center management. Key issues include; what degrees of variation, range or margin of error are required to provide an 80 to 90 percent reliable prediction of work for any period and what is driving the deviations from that norm. The chart below illustrates a 52 week forecast versus actual view.

Forecast vs actual


Periodic presentations to senior officers are always beneficial. With a forecast that is reliable and understood discussions focus on strategies for improvements and lessen those about what will or might happen. In short an accurate and effective scheduling process allows an organization to focus on strategic issues and not the tactical ‘’what ifs’ in the call center.

Forecasts like this are powerful tools. Updating them regularly, comparing actual results to projections management, staff and supervisors quickly gain confidence that management has foresight and is anticipating issues and problems before they arise. This tool provides more value at less cost than any other in the call centers arsenal.

Around Wikia's network

Random Wiki