log in to your account

welcome back! please log in to continue.


Contact Center Analysis: A Closer Look

February 12, 2014

Last week we saw a blog post that caught our interest. It outlined three models typically used in contact centers to do what-if analyses: predictive, descriptive and prescriptive. For each, there also were three tips for leveraging the model. Here is a quick synopsis.


Three models are commonly used by contact centers to perform what-if analyses.

Model 1: Predictive—forecasts call volumes, handle times, agent shrinkage and agent attrition

  • Forecast everything—including attrition and sick time.
  • Measure variance—the only way to improve forecasts.
  • Take warning—consistent variance from forecast is a sure sign of trouble.

Model 2: Descriptive—simulates the contact center environment

  • Ditch Erlang C—does not work with multi-skilled agents.
  • Validate—make sure your simulation makes real-world sense
  • Simulate discrete events—works well for multi-skill, multi-channel and multi-site centers.

Model 3: Prescriptive—recommends an optimal solution

  • Use integer programming—creates JIT plan at least cost.
  • Use integer programming scheduling—creates most efficiency.
  • Use for hiring plan—save your organization as much as 5-10% of paid agent hours.

Models can be terrific tools to improve efficiency when properly developed and validated. For the full blog post, visit “A Closer Look at Three Models Used for Contact Center Analysis“.