Skip to main content

Article 2 min read

Customize your forecast analysis period

You can now tell your forecast to ignore data before a specific date that you define by workstream to remove outliers and up your forecast accuracy.

By Lilith Shoemaker, Contributing Writer

Last updated May 3, 2024

Person looking at abstract screen

Accuracy for our forecast is paramount when we are planning our schedules and staffing needs, which is why outliers in past volume data can be a serious pain. If your business changes, causing a massive shift in volume, your forecast will likely be skewed depending on the previous data being considered. This is why you can now define which data Zendesk WFM uses when generating the forecast.

You now have the flexibility to tell your forecast to ignore data before a specific date that you define by workstream. This means a more accurate forecast for you, better coverage for your customers, and more favorable schedules for your agents.

Why this is valuable:

Say in the first year, your teams had a very low volume because you were not on your customer’s radar; however, this year, you had a business change that increased your volume immensely–maybe you launched a new viral product or got a jump start from a successful campaign. You would not want to include the first year in your forecast because the transition from the first to the second year is not an accurate representation of how your volume will grow in the future.

If we were using data from years one and two, your forecast would increase the volume drastically for year three rather than reflecting a forecast that leveled out in year two after the business change. Not only is this not ideal, but it’s usually a massive headache for the workforce team. That’s why Zendesk WFM is here to remove the issue before it occurs with a simple solution to eliminate the irrelevant outliers. In this case, we’d eliminate year one and use the data only after the growth.

Two things to keep in mind:

  1. You won’t be able to set this date back further than two years because Zendesk WFM is only looking at the past two years of data anyway.

  2. You also won’t be able to set a date closer than one month from when you are adjusting the data. Zendesk WFM requires at least one month’s worth of data for the algorithm to generate a forecast for you.

Give it a try:

  • Head into the forecast feature

  • Click the three dots next to the workstream you want to adjust

  • Select the “Historical volume” option

  • Click “Select Zendesk historical data”

  • It defaults to “Use all available data” so click to toggle that off

  • Select the specific date you would like and Zendesk WFM will ignore the data before that date

  • Click “Apply” and “Generate the forecast” to begin the forecast generation

Soon, you’ll have a forecast that more accurately reflects your needs based on the restrictions you determined were more reliable.

Related stories

Article
1 min read

AI for customer success: Benefits + use cases

Use AI for customer success to create personalized journeys, reduce customer churn, and lower business costs, all while improving customer experiences.

Article
1 min read

How to calibrate your customer service QA reviews

Here’s how to calibrate your customer service QA reviews to make sure all of your reviewers are aligned in providing consistent, helpful feedback.

Article
14 min read

Customer service ROI: How to measure and improve it

Customer service ROI is the total impact of investments in improving support quality and efficiency. Learn how to calculate and improve ROI in this guide.

Article
2 min read

Auto QA: How to automate customer service quality assurance

Automated quality assurance will help define the future of support. Find out what that means, and what Auto QA can do for you.