What's my plan?
Add-on Workforce Management (WFM) or Workforce Engagement Management (WEM)

Verified AI summary ◀▼

Learn how to set up workforce management forecasting to predict future contact volumes and staffing needs. Configure historical volume periods, choose forecasting algorithms, adjust for upcoming events, exclude outliers, and edit staffing parameters to optimize resource allocation and meet service goals. This setup helps ensure you have the right number of agents scheduled at the right times.

Forecasting in Zendesk Workforce management (WFM) is the process of predicting future contact volumes and staffing needs using historical data and advanced algorithms to ensure that the right number of agents are scheduled at the right times. This helps you optimize resources and consistently meet service goals.

In this article, you'll learn how to set up and manage your WFM forecasting by setting the historical volume period for each of your workstreams, selecting a forecasting algorithm, adjusting your expected volume based on unique upcoming events, excluding outliers to remove dates and times that may skew your forecast, and configuring your staffing parameters to calculate how many agents you should expect to schedule.

This article contains the following topics:
  • Setting the historical volume period
  • Selecting a forecasting algorithm
  • Adjusting the inbound volume
  • Excluding historical inbound volume outliers
  • Editing staffing parameters

Related articles:

  • Accessing and reading the WFM forecast
  • Editing and managing the WFM forecast page

Setting the historical volume period

For each workstream, you can set the historical volume period. This allows you to decide which historical data you want to include or exclude in your forecast calculation.

See About WFM forecasting to learn more about how your historical Zendesk ticket data is migrated.

To select the historical volume period
  1. In Zendesk WFM, hover over the forecast icon () in the navigation bar, then select Forecast.
  2. Select a workstream.
  3. In the Historical volume panel, select if you want to use your Zendesk data or to use Imported data as your Data source.

    • If you selected Zendesk data, choose whether to Use all available data from the last 2 years or to Ignore data prior to a specific date. Use the date picker to select your desired date.
      Important: The forecasting engine can use up to two years of data if it's available within WFM. Keep in mind that only six months of your historical ticket data from Zendesk is migrated into WFM when it's first activated for your account. You must contact Zendesk Support to manually import a longer historical period.
    • If you selected Imported data, download and use the Import volume template. See Importing historical volume for WFM forecasts.

  4. Click Save.

Selecting a forecasting algorithm

For each workstream, you can set the Forecasting algorithm.

The best forecasting algorithm for each workstream is automatically selected by analyzing your historical data and testing all algorithms. This is to determine which will perform best for the specific workstream. However, you can overwrite the automatic algorithm selection and choose any algorithm Zendesk WFM offers. See About WFM forecast algorithms to learn more.

To select the forecasting algorithm
  1. In Zendesk WFM, hover over the forecast icon () in the navigation bar, then select Forecast.
  2. Select a workstream.
  3. In the Algorithm panel, select the algorithm you want to use as Forecasting algorithm for this workstream.

  4. Click Save.

Adjusting the inbound volume

You can adjust each workstream's expected inbound volume based on unique upcoming events, such as marketing campaigns and feature removals.

To adjust your inbound volume

  1. In Zendesk WFM, hover over the forecast icon () in the navigation bar, then select Forecast.
  2. Select a workstream.
  3. In the Volume adjustments panel, click the plus icon (+) to add an adjustment.

    Here, you can:

    • Rename your adjustment to give it more context. For example, you can name it Marketing, New feature release, etc.
    • Select the date range to apply your volume changes.
    • Enter your expected inbound ticket volume under Change and select if this value is a percentage (%) or exact value (#).

  4. You can continue creating additional volume adjustments by clicking the plus icon (+).
  5. When you're done, click Save.

    Your inbound volume chart and required staffing counts are updated.

Understanding inbound volume adjustment distribution

The distribution of your inbound ticket volume adjustment is calculated rounding to one decimal place (for example, 10.1 instead of 10.123). Values are distributed in 15-minute intervals proportionally to what was calculated by the forecast. Any remainders are then distributed randomly, also following the patterns of the forecast.

The following examples illustrate different scenarios for this calculation.

Volume adjustment by number of tickets

Selecting changes by number of tickets for a specific time period distributes them according to the forecasted trend for the period, rounded to the first decimal.

For example, you expect 10 additional tickets on a future day. For the sake of simplicity, assume the forecast for this day was equal to all intervals, so tickets are equally distributed throughout the day.

The distribution of inbound volume adjustments is calculated as follows:

  • Each day consists of 96 15-minute intervals.
  • Dividing the +10 tickets by 96 equals 0.10416667. That’s then rounded to the first decimal, 0.1, and added to each 15 min interval.
  • The remaining 0.4 tickets (96 * 0.1 = 9.6 tickets, so 0.4 remain out of the total of 10) are distributed randomly, following the forecast pattern.

Volume adjustment by percentage

When you create an inbound volume that’s adjusted by percentage, the algorithm takes the absolute value of tickets for the selected period and applies the percentage to it. This results in +/- X number of tickets which is then distributed as described above.

For example, take the period of one day and set an increase of 100% for that specific day. The expected value was 20 tickets; with the 100% increase, it’s expected to double. The distribution process then occurs as described above.

Excluding historical inbound volume outliers

You can exclude outliers by removing dates and times that may skew your forecast. This can be useful for events such as a one-time marketing campaign or an outage that could have caused a sudden increase in your inbound volume.

After you've saved the excluded outliers, the forecast recalculates and displays without the impact of historical spikes. It may take a few moments for the forecast to recalculate

Note: This option isn't available for imported data. You must edit outliers within the import itself.

To exclude historical inbound volume outliers

  1. In Zendesk WFM, hover over the forecast icon () in the navigation bar, then select Forecast.
  2. Select a workstream.
  3. In the Excluded Outliers panel, click the plus icon (+) to add an outlier.
  4. Click in the name field or select the options menu icon (), then select Rename.

    Accurately rename the outlier. For example, Outage.

  5. Choose a start and end time and date as well as the time you want to exclude.
    If the exclusion date range selected doesn't have enough data preceding it to remove the outlier (a minimum of two weeks is required), this specific outlier won't be replaced and a warning will be displayed. The forecast will still be generated successfully.
  6. Click Save.
  7. If you want to delete an excluded outlier, click the options menu icon (), then select Delete. If you want to delete all of the outliers created, click the button at the bottom of the column.

Editing staffing parameters

The forecast allows you to modify the full-time equivalent (FTE) count parameters according to your set business requirements because not every workstream can have the exact first response time (FRT), average handle hime (AHT) and shrinkage expectations. Learn more about Required staffing.

You can also set specific business hours and a channel preference for a workstream. The channel preference controls staffing parameters that will be used to build your staffing forecast.

To modify your FTE count parameters

  1. In Zendesk WFM, hover over the forecast icon () in the navigation bar, then select Forecast.
  2. Select a workstream from the Forecast page.
  3. In the Excluded Outliers panel, click the arrow to expand the panel. From here you can:
    • Select the Channel preference for the workstream.
    • Adjust the Occupancy Rate for the workstream.
    • Change the AHT and select values for hh:mm:ss.
    • Specify the Target of the FRT and select the values for hh:mm:ss.
    • State the SLA attainment Target.
    • Enter a Shrinkage value.
    • Note: Shrinkage is the % value of unscheduled or scheduled activities (vacations, sick leaves, etc.) which prevent employees from performing their work. The industry standard is approximately 15-35%.
    • Add the minimum staffing number you need for that workstream.
    • Choose whether or not you want to specify your availability or business hours.
    • You can also reset to the default values.

  4. When you're done, click Save.
Currently, your minimum staffing number will be respected whenever possible when scheduling automatically. There may be instances, however, where high forecasted demand is prioritized over minimum staffing. If you've signed up for the Combined Workstream EAP and the workstream is associated with a combined workstream, the minimum number to be respected is the highest among all workstreams.
Powered by Zendesk