Skip to content

Predicting staffing efficiency at an inbound phone help desk in R.

Notifications You must be signed in to change notification settings

Drev917/Help-Desk-EDA

Repository files navigation

Help-Desk-EDA

Problem statement:

Examine average daily call patterns and staffing efficiencies during a representative population in the sample dataset.

ScreenShot

The Help Desk data table provides information on typical volumes for a week of help desk calls; This sample in particular was pulled from calls recorded in the month of October.

ScreenShot

Structure of the data is 60 observations of 3 columns:

  • Day & Hour vectors have character data type
  • Number of Calls is a numerical data type A quick summary of the data shows a total range of observable calls of 1,156, median value of 341, and a mean of 339 calls rounded as an integer.

Looking at the data table comprehensively, we quickly noticed a few outliers and values that we determined as erroneous inputs. We recoded negatives integers and erroneous character variables.

ScreenShot

Sliced the data by the unique Day values, added the Hour values as columns, and then stacked the calls horizontally, allowing us to easily see all the calls using the head() function. Used dplyr library to quickly sum by Number of Calls.

ScreenShot

Our team decided to use the mean statistic for the data set as the best metric to evaluate the data.

  • By using mean you can quickly calculate any day’s variance based on the sample size and staff accordingly
  • The mean also allowed us to create quick ‘Mean calls by day’ and ‘Mean calls by shift’ variables in R

ScreenShot

We can then present our major findings below:

ScreenShot

And our recommendations for action:

ScreenShot

Thank you for reading!

ScreenShot

Authored by Drew Barton, Chris Calvo, and Jake Pagels

About

Predicting staffing efficiency at an inbound phone help desk in R.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages