One of the biggest challenges in contact centers today is getting the forecasting and planning right. If you do that well, the rest is much easier. This 3-part series will help you incorporate some of the best techniques in the industry to forecast and plan smarter. We hope these tips and tricks help turn you into a forecasting hero!
No time to read the full article? Just download it as a PDF and read it offline, anytime, anywhere.
This article is the first of a 3-part series on Contact Center Forecasting Fundamentals. Keep on reading if you like to learn about the fundamental principles of contact center forecasting and dive deeper into the following topics:
In every article, we will provide helpful tips, expert advice as well as a set of strategies and tactics you can employ to get your forecast right and optimize your workforce management operations, one step at a time.
Contact Center Forecasting Methods and Tips
For anyone just getting started in forecasting for contact centers, it can be a real challenge. As you build out forecasting and planning, the expectations start to grow. People want more.
Whether it’s a stricter forecast accuracy target or moving to measuring at smaller time periods (e.g. from monthly accuracy to daily accuracy), as your operations mature the expectations will grow.
Let’s walk through some tips for getting started in contact center forecasting. The first tip, regardless where you are on your journey is to have a clear objective. It’s never forecasting just for forecasting's sake.
Download the free E-Book
The Must-Have Guide to Accurate Call Center Forecasting
What you will learn:
- Foundations of Call Center Forecasting
- Getting your data right
- The importance of measuring variability
- Critical factors you shouldn't miss
- And much more!
Long-Term Monthly Forecasting
The objective of monthly forecasting is to give the operations “directionally correct” information. What does that mean?
Well, it means you don’t spend a year trying to build a super-accurate forecast. You can get there over time. Instead, you focus on what you can, build quickly and provide some insights into how many Full Time Equivalents (FTEs) you need compared to how many FTEs you have or plan to have.
An FTE is the equivalent of one person working full time. The actual number of hours may vary by region. In the US, an FTE is 40 hours per week. The gap between what you have and need is what the operations needs to solve for.
Your monthly forecast should drive your hiring and staffing plan. This is the goal of this process. If your monthly forecast doesn’t ultimately position you to feed into a hiring plan, then it’s missing the mark.
To determine how many FTEs you require, you’ll need to forecast the workload. Workload is simply the contact or call volume (the number of incoming messages or calls) multiplied by the average handle time (AHT) of a call. Average handle time is the average time needed for a call including hold times and after-call work.
You need to build forecasts for volume and handle time separately. Both of these forecasts will start with historical data which your center has collected over the past months or years. After creating a forecast based on historical data, you layer in business intelligence, i.e. information about future events that might influence your forecast like e.g. a marketing campaign or a product release.
After you have created both the call volume forecast and the handle time forecast, you multiply call volume with the handle time to derive your workload forecast. The workload forecast is then used to calculate the needed staff per time period.
Let's delve into the details of the call volume forecast.
Forecasting Call Volume
Start by selecting a staff group to forecast. Gather as much historical data as you can for that group. Ideally, you go back 36 months, so you have a good view of how volume has trended over time. By that, you also get a sense of seasonality.
The easiest way to start is to identify the growth year-over-year to establish where your volume should be this year. Here is a sample data set for us to apply a simple growth factor to:
Table 1: YoY growth in contact volume
Since your growth rate was 4% one year and 3.8% the following year, the simplest approach is to just take the average growth rate to create the forecast. In this example, the average is 3.9%, so you apply it to forecast the next year.
With this growth rate, you would be adding 2,106 to the volume you saw in 2018. This gets us to a forecast of 56,106 contacts for 2019.
To shape this into a monthly forecast, calculate the average volume per month. Then add up all averages from all months to get to the average annual volume. In our case, this is 52,000. Based on that average annual volume and the average volume per month, determine what % of the year’s contacts come in each month on average. You do that by dividing the average monthly volume by the average annual volume for each month.
Table 2: Average contact volume per month based on sample data
For the 2019 forecast, you then just multiply the percent for each month by the total volume for the year.
Assuming your full year forecast is 56,106, your monthly volumes would look like this:
Table 3: Forecasted contact volume per month based on sample data
So this is a simple way to quickly get a call volume forecast that will provide some high-level direction to your operations.
So far we have only created forecasts for call volume. In order to determine the number of staff needed for each month which is our ultimate goal, we need to forecast average handle time (AHT) as well.
The AHT forecast is basically created the same way as the call volume forecast. You determine the annual AHT growth rate and calculate the AHT for the upcoming year. Then you calculate the monthly averages for the AHT and compare them to the annual average AHT to determine the deviation in percent. Then you use these monthly percentages together with the already calculated AHT for the upcoming year to determine the monthly AHTs for the year.
Creating the Workload Forecast
After you have calculated the AHT for all months, you multiply the AHT number for every month with the call volume of each month. The result is the workload for each month of the next year.
You can take the same process that we just followed to arrive at a weekly long-term forecast which might be helpful for some long-term staffing decisions.
Short-Term Weekly and Daily Forecasting
Weekly and daily forecasts are typically done 3-6 weeks in advance. The objective of these short-term forecasts is to determine where you schedule your staff in the following weeks. One question often asked is: Do short-term weekly and daily forecasts have to exactly match the monthly forecasts from long-term planning?
The answer is no. Your monthly forecasts are done farther in advance with less accurate information. As you get closer to the date, you should have updated business intelligence to help improve the accuracy of your forecast. Remember: Business intelligence, according to our definition, is all information about future events that might influence your forecast like e.g. a marketing campaign or a product release.
So if your short-term weekly and daily forecasts are simply a mathematical breakout of the monthly long-term forecast, you are missing some opportunities.
Let’s say you’re looking 3 months out. You have created a monthly forecast. You may also have created a weekly forecast for that same timeframe. This far out, it makes sense for them to match. But as you get closer, like 3-6 weeks, you should adjust your weekly forecast which was based on long-term data. You create a new short-term weekly forecast that helps you to fine-tune staffing.
As you have new information to change your weekly forecast, you should document that change and give the reasons why.
Here is an example: Your forecast 3 months from now assumes your call volume will grow 10% year-over-year, which is consistent with how you’ve been performing.
When you get a month out, you learn there is a change in the billing process that will likely result in customers calling in to ask questions. In this case, you will want to increase the volumes in your weekly forecast and document the reason for the change. By documenting, not only do you make it clear why a change was made, but you can also then test your assumptions against the actual results.
If you thought the extra billing calls would drive an added 5% in volume and compare that to what actually happened, you can use this the next time there is a billing change to have an even more accurate forecast.
Let’s now take a look at two common methods to calculate short-term daily forecasts.
You basically take the same approach as in long-term planning.
- You first calculate the call volume for the whole year based on growth rates of the previous years.
- Then you calculate the average call volume for each day of the year that is of interest for you.
- After that, you divide the call volume of each day by the average annual call volume to determine the percentage of each day.
- Lastly, you multiply the percentage of each day with the call volume of the current year that you already calculated. As a result, you’re getting the call volumes for each day.
After you’ve applied these steps, you need to adjust for holidays.
Some holidays fall on the same date (like Christmas). Others may fall on a Monday, so the date will change each year. To adjust your holidays, look at the day of week distribution for the week before, during and after the holiday last year.
You want to see how the call volume each day differs from that of normal weeks. You calculate the percentage difference for each day and apply that to the daily call volumes for the weeks around each holiday. In many cases, you may only have volume impacted on the actual week of the holiday. If that’s the case, you only need to adjust that week’s data.
I recommend the first time you do the holiday factors that you look at all 3 weeks, then determine which days are actually affected. Save the change percentages for all affected days and you can use them again in future years.
>> Click here to uncover if your contact center is lacking forecast accuracy.
When using this method, you create the monthly forecast as we did in long-term forecasting. Once you have the monthly volumes, you use the historical distribution of volume-by-day for the month.
For example, if Jan 1st is generally 2% of your January volume, and your January forecast is for 10,000 calls, then your forecast for January 1st is 200 calls.
Of course, you need to adjust for holidays as we did in Method 1. This is an important step that can easily get overlooked if you don’t have a checklist to make sure you know all of the holidays that influence call volume distributions.
For those who take contacts from different markets, keep in mind your callers may live in a country (or state) that has different holidays than your contact center taking the calls. You’ll need to do some research.
Forecasting AHT and Creating the Workload Forecast
After creating your call volume forecasts, it’s time again to create the same for AHT. The procedure is the same. Once you have the forecasts both for call volume and AHT in place, you can combine them to a workload forecast by multiplying call volumes with AHT values for each day.
Business Intelligence in Forecasting
I can’t state this emphatically enough. You need to layer in business intelligence over any data-driven forecast, be it long-term or short-term. This is true both for call volume as well as for AHT forecasts. Business Intelligence is any information you can get that explains why the future will be different than the past.
The billing process example that was used above is pure Business Intelligence. Other examples would be marketing campaigns, new product launches, changes in policies, or changes to online self-service. There are dozens of things that can change going forward. You need to be in the loop on all of them.
A best practice here is to set up weekly and monthly meetings with various departments that may impact volume. Set up a strategic monthly meeting and a more tactical weekly meeting.
In the monthly strategic meeting, you want groups like marketing, sales, operations, billing, etc. to participate. In the session, you should cover a list of things you know have impacted volumes in the past to see if they have any similar in the future and ask them if there is anything else that they know of that may impact volumes.
The reason why it’s good to come in with a list of potential impact areas, is that it helps ensure you’re not just relying on what they think may impact volume.
In your weekly tactical meeting, it’s best to recap the expected impacts from the monthly meeting and see if anything has changed from your previous assumptions. Additionally, you’ll want to cover anything shorter term that may impact volumes. One example here is recontact volume. These are the callers that abandoned their calls due to long wait times and will likely call again in the following days. You have to make sure that this additional contact volume is added to your short-term forecast.
>> Learn more about how to incorporate Business Intelligence in your forecasting.
Pro Tip: Boosting forecast quality and efficiency with WFM software
The forecasting steps in this article can be applied regardless of your WFM setup. As some of the calculations are time consuming, a spreadsheet tool like Excel can be an improvement. Spreadsheets will help you automate many steps. They will also reduce the likelihood of calculation errors.
However, spreadsheets tend to get complex, hard to maintain and error-prone over time. Learn more about the disadvantages of Excel in this article.
If you are serious about improving the forecast quality and efficiency of your contact center, you may want to consider trying out a professional WFM software.
Advantages of WFM tools in terms of forecasting
- Less time and effort: Modern WFM tools create forecasts very quickly. All you have to do is plug in a source for the historical data and the forecast will be created for you.
- Higher forecast accuracy: WFM tools, especially those powered by AI like injixo, create forecasts that are often much more precise than manual or semi-automated calculations.
- Always up-to-date: Whenever new data becomes available, WFM software can do an automatic re-forecast. This results in higher forecast accuracy. Re-forecasts can also be triggered throughout the day to enable schedule changes for real-time management.
- Business intelligence built-in: Anomalies in historical data called “events” can be spotted and linked to certain event types like a marketing campaign or a billing process change. Whenever the event occurs again in the future, the forecasts will automatically take that into account.
>> If you want to learn more about how WFM software can improve forecasting in you contact center, feel free to contact the WFM experts at injixo.
Now that you’ve created a workload forecast, the next step is to forecast your staffing.
It’s not as simple as just taking today’s staff and apply attrition. There is much more to it than that and we’ll cover that in our next blog.
>> Proceed with #2: How to Master Workforce Forecasting
Did you find the article interesting and would like to share it with your colleagues? Download the article as a PDF.
Originally published on May 29, 2018 , updated on May 12, 2020