Contact Center Operations
Dec 1, 2022
Contact center forecasting determines staffing needs for specific intervals of time, like days, weeks or months. Yet, only 20% of service professionals state that their organizations excelled at forecasting demand in a Salesforce State of Service report.
Contact center forecasting ensures the right number of agents are scheduled for the expected support volume across multiple channels such as chat, email, social, and phone.
This article highlights methods, benefits, best practices, and factors to consider in an effective forecasting and scheduling solution.
Table of Contents
Spreadsheets as a workforce forecasting and scheduling tool should be a thing of the past
An accurate forecast is only effective when the corresponding schedule meets and dynamically adjusts to customer demands. However, many organizations use spreadsheets to create, manage, and maintain schedules, attendance, and staffing forecasts. This manual process is not only a time-consuming and laborious, it also inhibits the ability to quickly make and communicate changes in realtime.
A recent survey conducted on contact centers reveals that 50% of call centers manually create forecasts for digital channels, lowering planning efficiency and effectiveness. Workforce managers and support leaders need to easily visualize forecasted schedules across multiple channels to easily understand where the team is understaffed, overstaffed, and covered, and instantly make adjustments to meet demand.
A more comprehensive workforce management solution that automates schedule creation and forecasts based on historical demand, seasonality, and performance data to meet target SLAs, response times, and productivity levels is vital for success.
Understanding call center forecasting models
Effective and efficient scheduling relies on good forecasting. Workforce forecasting predicts future headcount needs for ongoing or upcoming initiatives such as new product or service launch, marketing promotions, seasonal events, and more.
Workload projections guide workforce forecasting, and uses historical trends, performance or seasonal data to predict personnel needs. To forecast accurately, team leaders need to adopt an effective technique.
Here are five leading call center forecasting methods:
1. Triple Exponential Smoothing
Known as the Holt Winters technique, this forecasting method splits forecast data into three components — level, trend, and seasonality — and averages the inputs from one period to the next.
For example, in a monthly forecast, the three components translate to:
Level. The last month’s forecast.
Trend. The increase or decrease in anticipated contacts from the previous month.
Seasonal. The effect of seasons on data. This component measures the difference between the general average and a specific month.
This forecast model is easy, but teams need to be mindful not to “overfit” data to avoid historical volume anomalies such as demand peaks or outages, as this can lead to an odd-looking forecast.
2. Autoregressive Integrated Moving Average (ARIMA)
A slightly advanced call center forecast method, ARIMA encompasses three main areas:
Auto-Regression. This involves comparing data against past patterns, such as results from 52 weeks ago.
Integrated. The difference between past and present observations.
Moving average. The focus here is to smooth out data over given periods.
ARIMA leverages historical data to present data sets based on past inputs. Data from past periods smooth out existing inputs and make forecasts more accurate.
3. Neural Networks
Leading organizations have adopted neural networks for artificial intelligence in areas such as search algorithms and speech analytics. This forecasting model tries to model the brain by observing a series of inputs and then attempting to adjust a “hidden network” until it discovers a matching output. In a call center, a neural network will examine a series of calls and attempt to match the next field of data to the forecast.
Neural networks are flexible as they can accommodate external inputs such as special days, website views, and marketing activity. Also, this forecasting model does not require complex algorithms as it learns and improves from existing data, automatically isolate specific days from a forecast, and model different factors.
Neural networks are input heavy, time-consuming, and may not be best for teams that rely heavily on trends.
4. Multiple Temporal Aggregation (MTA)
This forecasting method combines high-frequency data — hourly and daily inputs — with trends that span an extended period. An example of MTA in practice will be comparing the number of contacts acquired in 2021 with 2020 and getting an 8% increase.
The result, which is the trend, averages out the contacts and special events, such as seasonality over the year. MTA allows teams to focus on intraday and longer-term data in generating forecasts.
5. Erlang C Formula (industry standard headcount prediction)
At its core, Erlang C allows you to model the relationship between staffing, support volume, and response time. Conceived by Danish mathematician Agner Erlang in 1917, Erlang C is a capacity planning tool that allows workforce managers to identify their staffing needs by inputting in the number of agents they have at their disposal, their support volume, and the average response time of their operation.
Erlang C analysis removes some of the guesswork by clarifying how staffing decisions impact bottom-line customer-facing outcomes such as response time and service level, and ensures the ideal number of agents are available to meet demand every time.
Echo AI Workforce Management combines Erlang C with proprietary algorithms to predict future ticket volume to accurately forecast future headcount needs to meet target SLAs, response times, productivity levels, and more, across multiple channels.
Benefits of contact center volume forecasting
Contact center forecasting projects the number of agents required to meet future demand. It improves both customer and agent experience, and puts the team in a position to succeed.
Eliminate costly wait times: With 64% of consumers saying they would never shop with a retailer again if they abandoned a customer service conversation before being helped due to wait times, workforce forecasting ensures the right number of agents, with the right skillsets, are scheduled at the right time to meet customer expectations and improve customer experience.
Avoid agent burnout: Even before the pandemic, 74% of call center agents were at risk of burnout. Accurate forecasts enable efficient and fair distribution of workload to eliminate agent burnout and improve job satisfaction.
Make well-informed hiring plans: With staff costs typically making up approximately 70% of a contact center’s total costs, forecasting offers visibility and clarity to make informed staffing decisions and optimize spend.
Improve employee morale: According to a recent SWPP survey, forecast accuracy is actually the #1 measure that affects team satisfaction. Accurate workforce forecasting is vital for maintaining employee morale.
Contact center forecasting and scheduling best practices
Factors to consider in a workforce management tool
A workforce management tool such as Echo AI aids in forecasting success and facilitates schedule adjustments based on changes in demand levels.
Here are some other features to examine when selecting a call center forecasting tool:
Supports forecasting for multiple channels, activities, and time zones
Offers an easy-to-use interface to create, modify, and publish schedules
Leverages historical demand, seasonality, and performance data to predict future headcount needs
Provides tools to manage attendance and approval processes, and integrates with your existing payroll systems to streamline operations
Allows you to create and customize events such as breaks, lunches, holidays and more
Integrates with your existing support platform to track and measure agent activity
Offers robust reporting and analytics to understand schedule efficacy and adherence at channel and activity level
Choose Echo AI as your workforce management solution
The ability to use performance data in workforce forecasting is crucial. Manual prediction and scheduling is time-consuming, laborious, and error-prone, and does not offer the flexibility to make realtime adjustments to meet demand. While historical volume can predict future demand, integrating performance and productivity into the calculation can make a big difference.
Using a dedicated workforce management platform, service leaders and managers can stay on top of support volume and meet customer expectations with data-driven forecasting and scheduling.
Echo AI for Workforce Management simplifies scheduling, forecasting and attendance processes to help businesses stay on top of support volume and meet demand at scale.
Learn more about how Echo AI can optimize WFM in your business. Request a demo today.
