First of all, why does contact centre forecast accuracy matter? In my experience, the most important outcome from a forecast is the headcount or FTE (Full Time Equivalents) that you need. Every other sub-forecast (contacts, handle time, shrinkage) are all critical to contributing to the headcount forecast.
Forecasting is a function that ranges from hyper-technical to very basic. Neither extreme is helpful. For those who take only a statistical approach, a lot of important factors are missed, and often it’s challenging for business leaders to use the output for decisioning. Taking a basic approach and doing very simple calculations is often a result of lack of tools or resources to generate more sound forecasts.
Ultimately, you want to be closer to the middle. The base of your forecast should leverage historical data to project forward, and statistics are very helpful here. The outputs need to be modified using business intelligence to improve accuracy. And ultimately, your output needs to be simplified so leadership can understand the conclusions and scenarios and make a decision.
In this article, Charles Watson of Injixo covers what can hurt your contact centre forecast accuracy.
1. Forecasting Contact Volume Without Business Intelligence
As I mentioned above, this is an important modifier. When using regression analysis to project forward, you assume what’s happened in the past will continue into the future. This is valid in some forecasting, but not in contact centres.
There are too many human decisions and behaviours that are fluid. Historical data will only get you so far. It’s equally important that you get insights into how things in the environment will be different in the future to adjust the statistical forecast.
Where contact centre forecasting on historical data doesn’t work:
A common example is a marketing campaign. If your business is going to conduct a marketing campaign, it is going to take an active approach to generating more revenue from your customers.
The campaign may increase contact rate or drive incremental contacts into a small staff group. Other information that helps improve contact centre forecast accuracy is:
- Changes in the operations: changes in how agent time is focused
- External environment: changes in consumer confidence, changes in unemployment
- Consumer behaviour: changes in what types of products they buy, changes in brand reputation
2. Forecasting Handle Time That Is Only Based on the Average
To determine workload, the expected volume is multiplied by the average handle time. That tells you how many total hours of work actually need to get done.
Because of this, the work on forecasting handle time generally takes the historical average handle time and projects it forward.
This projection needs to include expected learning curves for new hires. But there is even more beneath the surface here.
Is your average a tight distribution across the agents or is there wide variation?
If you have a tight distribution (a small standard deviation), then this number is more reliable. But if you have a wide variation, then the number may be driven by a small number of agents.
If those agents leave, get better or worse in their performance, they can change this number significantly. Take the extra step to see the spread of handle time across the population. If you see some outliers, then you need to track those agents separately to see how they are trending.
This is also a great opportunity to better understand from their leaders if they expect performance to change.
It’s also a good practice to understand whether there are any expected process changes. Handle times are heavily impacted by how simple or complex the process are that the agents need to execute. If a step gets added or removed, you should see a change in handle time.
3. Forecasting Attrition Without Talking to HR and Operations
A reality of staff planning is the movement of people across various queues as well as into and out of the organisation.
With all the work that goes into determining the headcount required, it’s important to make sure you have a good idea of exactly how many people you’ll actually have to meet that requirement.
There are two types of attrition: voluntary and involuntary.
a. Voluntary attrition in your contact centre
This is when agents leave your organisation on their own. This may be driven by low employee morale or satisfaction. It can also be driven by new or changing competition for talent.
Understand how employee satisfaction data can give you insights into a potential increase in attrition. And understand the local job marketplace so you know where there may be risk. HR should have insights into both of these that you can use.
b. Involuntary attrition in your contact centre
This is when agents are terminated for performance or other reasons. Performance management is the effort within the operations to improve agent performance in several different areas important to the organisation.
When companies decide to performance-manage more aggressively than in the past, it can result in higher termination rates, which impacts involuntary attrition.
4. Forecasting to a Budget Instead of Reality
Those of you who have been doing contact centre forecasting for any significant amount of time and are successful forecasters, you’ve probably been asked to forecast to a budget number instead of what you’re actually seeing.
This happens because the business leaders want to set public targets for their teams and have a clear view of how performance is trending against this target.
Unfortunately, you lose sight of what’s actually needed for service level.
An example of this is a business target of a 9-minute handle time that you are required to reflect in your forecast. The actual trend is 11 minutes. In putting in 9 minutes, the forecast will show you need fewer people than you actually do.
If you’re not in a position to change the methodology of having budget in the forecast, then you should show both the budget and the trend at the same time.
Yes, it’s more work, but it’s critical to show the difference between the desired state and reality. It also ensures you always have a clean view of what’s needed for service level and make sure your forecast is not missing any critical factor.
5. Not Forecasting for Occupancy
Many people think of occupancy as an output of the forecasting process, and they are right. However, it’s also an input into the process when you’re planning staff.
For any staff group, there will be a correlation between the occupancy and the service level. Simply using an Erlang formula when doing the monthly staff requirements won’t give you what you need.
Your occupancy input has a direct impact on the staff you need to achieve service level. And this relationship changes as other factors change.
Let’s explore a few of those.
A group has historically achieved an 85% occupancy when they hit an 80% service level. If the schedule efficiency changes because agents now get two consecutive days off, then your occupancy will go down.
You may now only get an 82% occupancy at an 80% service level. The drop in occupancy means you now need more people to achieve your service level.
Other factors that will change the relationship between Occupancy and Service Level
- Changes in hours of operation
- Consolidating groups
- Splitting apart groups
- Significant changes in handle time
- Increase or decrease in cross-utilisation across groups
Don’t Let These 5 Mistakes Ruin Your Contact Centre Forecast Accuracy
You spend a lot of time and effort on your contact centre forecast. Consider how all these factors impact the quality of your forecast.
Make a few simple changes, include regular conversations with marketing, HR, and operations to make sure you have all the information you need to deliver a world-class forecast!
This blog post has been re-published by kind permission of injixo – View the original post
To find out more about injixo, visit their website.