WFM Labs Intraday Performance Analysis
Intraday Staffing Efficiency Analysis is a critical tool for identifying and mitigating risks in contact centers by leveraging real-time data to optimize staffing and reduce costs. In today’s fast-moving contact centers, traditional staffing models—often dependent on rigid pre-planned schedules and manual adjustments—frequently fail to address unexpected changes in demand or agent availability. These outdated models, while built on detailed forecasts, can’t adapt quickly to the natural fluctuations that occur throughout the day, leaving contact centers vulnerable to service level failures or unnecessary expenses.
By adopting an Intraday Staffing Efficiency Analysis framework, workforce management (WFM) professionals can harness the power of Prior Day Performance (PDP) data and automation to rapidly assess key performance metrics. This framework enables WFM teams to evaluate real-time data streams, including forecasted vs. actual call volumes, staffing levels, and adherence metrics. With these insights, teams can dynamically adjust staffing, reallocate resources, or activate off-phone activities in response to emerging risks like service level drops or overstaffing, ensuring continuous operational efficiency
Intraday Performance Analysis Framework
Intraday data can vary depending on your WFM platform, but leveraging tools like Python and data notebooks allows for rapid, reusable code that can assess your unique data. In the following example, we analyzed a single day of interval-level forecasted and actual metrics, including call volume, AHT, scheduled staff, actual staff, and required staff. This analysis highlights the impact of over- and understaffing on your results.
Click "Run" to examine the results on this sample day's data.
Run to view results
Let's enhance the analysis by adding the following: 1. Calculate the total cost of overstaffing and understaffing. 2. Calculate the average duration of overstaffing and understaffing periods. 3. Identify the peak times for overstaffing and understaffing. Let's start by calculating the total cost of overstaffing and understaffing. For this, we need to assume some cost metrics: - Cost per overstaffed employee per interval. - Cost per understaffed employee per interval. Let's assume: - Overstaffing cost per employee per interval is $20. - Understaffing cost per employee per interval is $10. We'll add these calculations to the existing code.
Run to view results
The total cost of overstaffing is $1240, and the total cost of understaffing is $1420. Next, let's calculate the average duration of overstaffing and understaffing periods.
Run to view results
The average duration of overstaffing periods is approximately 4.33 intervals, and the average duration of understaffing periods is approximately 5.33 intervals. Finally, let's identify the peak times for overstaffing and understaffing.
Run to view results
The peak times for overstaffing are: - 10:30 AM to 12:30 PM - 5:30 PM to 6:30 PM - 8:00 PM to 10:00 PM The peak times for understaffing are: - 8:00 AM to 10:00 AM - 1:00 PM to 5:00 PM - 7:00 PM to 7:30 PM
what additional analysis can we perform?
Here are some additional analyses we can perform: 1. **Service Level Analysis**: Calculate the service level percentage for each interval and identify intervals where the service level falls below a certain threshold. 2. **Abandonment Rate Analysis**: Calculate the abandonment rate for each interval and identify intervals with high abandonment rates. 3. **Forecast Accuracy**: Analyze the accuracy of the forecasted calls and AHT compared to the actual values. 4. **Correlation Analysis**: Examine the correlation between different variables such as forecast accuracy, service level, and staffing levels. 5. **Trend Analysis**: Identify trends in staffing levels, service levels, and costs over time. 6. **Cost-Benefit Analysis**: Compare the costs of overstaffing and understaffing with the benefits of maintaining service levels. Let's start with the service level analysis. We'll calculate the service level percentage for each interval and identify intervals where the service level falls below a certain threshold (e.g., 80%).
Run to view results
The intervals where the service level falls below the 80% threshold are: - 8:00 AM - 10:00 AM - 1:00 PM to 1:30 PM - 2:30 PM to 3:00 PM - 4:30 PM to 5:00 PM - 7:30 PM Next, let's perform the abandonment rate analysis. We'll calculate the abandonment rate for each interval and identify intervals with high abandonment rates (e.g., above 5%).
Run to view results
The intervals where the abandonment rate is above the 5% threshold are: - 8:00 AM - 4:30 PM Next, let's analyze the forecast accuracy by comparing the forecasted calls and AHT to the actual values. We'll calculate the percentage deviation for both calls and AHT.
Run to view results
The average percentage deviation for forecasted calls is approximately 3.56%, and for AHT, it is approximately 3.73%. Next, let's perform a correlation analysis to examine the relationships between different variables such as forecast accuracy, service level, and staffing levels.
Run to view results
The correlation matrix shows the relationships between different variables: - There is a strong positive correlation between `Act Staff` and `Sched Open` (0.999) and between `Act Staff` and `Act Req` (0.980). - There is a moderate negative correlation between `Calls % Dev` and `Act Staff` (-0.460) and between `Calls % Dev` and `Sched Open` (-0.448). - The service level (`Svc Lvl`) has weak negative correlations with most variables. Next, let's perform a trend analysis to identify trends in staffing levels, service levels, and costs over time.
Run to view results
The trend analysis plots show the following: 1. **Staffing Levels Over Time**: The scheduled, required, and actual staffing levels throughout the day. 2. **Service Levels Over Time**: The service levels throughout the day, with a threshold line at 80%. 3. **Staffing Costs Over Time**: The overstaffing and understaffing costs throughout the day. Next, let's perform a cost-benefit analysis by comparing the costs of overstaffing and understaffing with the benefits of maintaining service levels.
Run to view results
The total service level benefit is 1882.7, but the net benefit, after subtracting the total costs of overstaffing and understaffing, is -777.3. This indicates that the costs outweigh the benefits in this scenario. If you need further analysis or have any specific requests, let me know!