Outbound Call Analysis

All Call Data

All calls placed from V4W were merged with a newly created case data (most recent in the case that we have a client in the dataset who previous had a case/account with us) from Health Cloud using the requester’s phone number. Then, account data associated with the case was imported as well. All test cases and duplicates were removed from the dataset.

V4W records their calls in EST, and health cloud data is stored in GMT. To account for the difference, 5 hours in time was added to each V4W call record.

The time required for V4W to initiate a phone call from a ROGER Help Request form was calculated by taking the date and time stamp of the V4W outreach and subtracting it from the created date and time of the Help Request submission. Callback times that were greater than 1000 minutes were excluded as outliers. In this case, that was 1 call record.




Discrepencies Depending on Day and Time




Graphics above display discrepancies in callback time that occur based on the day of the week and the hour of the day. Webform submissions are shown in blue, and V4W calls are shown in red.

Ideally, there would be an equal number of requests and placed calls daily. The exception would be when a request is placed within 15 minutes of midnight, and the callback occurs just after midnight the following day. When there are less calls placed in a day than requests, it indicates that there may be significant lapses in callback time. When there are more calls than requests, it indicates that calls are likely being made for requests that were placed the day prior.

When looking at the hour of day the calls are placed, the same can be evaluated. More incoming requests than outgoing calls within an hour may indicated prolonged callback waits at that time of day.

The table below shows summary data for time periods over the last three years. Summaries include:

  • Number of calls: Total count of web requests followed by phone outreach.
  • 25% Quartile: Time in which 25% of all web requests are responded to by phone outreach.
  • 50% Quartile (Median): Time in which 50% of all web requests are responded to by phone outreach.
  • 75% Quartile: Time in which 75% of all web requests are responded to by phone outreach.
  • Mean: Average time required to respond to web request by phone outreach.
  • Maximum: Longest recorded response time to a web request.
  • Standard Deviation: Indicates how much response times vary from the average.

The table can be sorted using each column. Additionally, you can search within each column or within the entire table using the “search” box. The table can be exported to CSV, Excel, PDF. It can also be copied or printed using the links in the top left hand corner.




Calls greater than 15 minutes

Here, we limit the analysis to all calls in which response time exceeded our goal of 15 minutes.

Graphics below display discrepancies in callback time that occur based on the day of the week and the hour of the day. Webform submissions are shown in blue, and V4W calls are shown in red.

Ideally, there would be an equal number of requests and placed calls daily. The exception would be when a request is placed within 15 minutes of midnight, and the callback occurs just after midnight the following day. When there are less calls placed in a day than requests, it indicates that there may be significant lapses in callback time. When there are more calls than requests, it indicates that calls are likely being made for requests that were placed the day prior.

When looking at the hour of day the calls are placed, the same can be evaluated. More incoming requests than outgoing calls within an hour may indicated prolonged callback waits at that time of day.




In the figure below, we show the distribution of outlier calls (those 25% that took the longest to receive a phone outreach). Here we display the proportion that received a call within 1 hour, between 1 and 2 hours, and between 2-4 hours.

We also show the distribution by day of the week to look for particular days (such as weekend outreaches) that may be effecting average response time.




Nighttime Data

This data is limited to the help request forms that are submitted between 12am and 6am EST.

The table below shows summary data for time periods over the last three years. Summaries include:

  • Number of calls: Total count of web requests followed by phone outreach.
  • 25% Quartile: Time in which 25% of all web requests are responded to by phone outreach.
  • 50% Quartile (Median): Time in which 50% of all web requests are responded to by phone outreach.
  • 75% Quartile: Time in which 75% of all web requests are responded to by phone outreach.
  • Mean: Average time required to respond to web request by phone outreach.
  • Maximum: Longest recorded response time to a web request.
  • Standard Deviation: Indicates how much response times vary from the average.

The table can be sorted using each column. Additionally, you can search within each column or within the entire table using the “search” box. The table can be exported to CSV, Excel, PDF. It can also be copied or printed using the links in the top left hand corner.




Outlier Analysis

The figures below represent only those outreaches that fell outside of the 75% quartile. In other words, these are the top 25% slowest responses.

Here we display discrepancies in callback time that occur based on the day of the week. Webform submissions are shown in blue, and V4W calls are shown in red.

Ideally, there would be an equal number of requests and placed calls daily. The exception would be when a request is placed within 15 minutes of midnight, and the callback occurs just after midnight the following day. When there are less calls placed in a day than requests, it indicates that there may be significant lapses in callback time. When there are more calls than requests, it indicates that calls are likely being made for requests that were placed the day prior.




In the figure below, we show the distribution of outlier calls (those 25% that took the longest to receive a phone outreach). Here we display the proportion that received a call within 1 hour, between 1 and 2 hours, and between 2-4 hours.

We also show the distribution by day of the week to look for particular days (such as weekend outreaches) that may be effecting average response time.




By Peer Statistics

Data is displayed by V4W peer. All data is displayed in minutes. The following data is summarized for each specialist:

  • Sample Size: Total count of web requests followed by phone outreach.
  • 25% Quartile: Time in which 25% of all web requests are responded to by phone outreach.
  • 50% Quartile (Median): Time in which 50% of all web requests are responded to by phone outreach.
  • 75% Quartile: Time in which 75% of all web requests are responded to by phone outreach.
  • Mean: Average time requested to receive a call back following a web request.
  • Maximum: Longest recorded response time to a web request.
  • Standard Deviation: Indicates how much response times vary from the average.

The table can be sorted using each column. Additionally, you can search within each column or within the entire table using the “search” box. The table can be exported to CSV, Excel, PDF. It can also be copied or printed using the links in the top left hand corner.







This data is presented again, including only nighttime requests occurring between 12am and 6am.

Data is displayed by V4W peer. All data is displayed in minutes. The following data is summarized for each specialist:

  • Sample Size: Total count of web requests followed by phone outreach.
  • 25% Quartile: Time in which 25% of all web requests are responded to by phone outreach.
  • 50% Quartile (Median): Time in which 50% of all web requests are responded to by phone outreach.
  • 75% Quartile: Time in which 75% of all web requests are responded to by phone outreach.
  • Mean: Average time requested to receive a call back following a web request.
  • Maximum: Longest recorded response time to a web request.
  • Standard Deviation: Indicates how much response times vary from the average.

The table can be sorted using each column. Additionally, you can search within each column or within the entire table using the “search” box. The table can be exported to CSV, Excel, PDF. It can also be copied or printed using the links in the top left hand corner.




Monthly Data

Data is displayed by month of help request submission. All data is displayed in minutes. The following data is summarized for each month:

  • Sample Size: Total count of web requests followed by phone outreach.
  • 25% Quartile: Time in which 25% of all web requests are responded to by phone outreach.
  • 50% Quartile (Median): Time in which 50% of all web requests are responded to by phone outreach.
  • 75% Quartile: Time in which 75% of all web requests are responded to by phone outreach.
  • Mean: Average time requested to receive a call back following a web request.
  • Maximum: Longest recorded response time to a web request.
  • Standard Deviation: Indicates how much response times vary from the average.

The table can be sorted using each column. Additionally, you can search within each column or within the entire table using the “search” box. The table can be exported to CSV, Excel, PDF. It can also be copied or printed using the links in the top left hand corner.




This data is presented again, including only nighttime requests occurring between 12am and 6am.

Data is displayed by month of help request submission. All data is displayed in minutes. The following data is summarized for each month:

  • Sample Size: Total count of web requests followed by phone outreach.
  • 25% Quartile: Time in which 25% of all web requests are responded to by phone outreach.
  • 50% Quartile (Median): Time in which 50% of all web requests are responded to by phone outreach.
  • 75% Quartile: Time in which 75% of all web requests are responded to by phone outreach.
  • Mean: Average time requested to receive a call back following a web request.
  • Maximum: Longest recorded response time to a web request.
  • Standard Deviation: Indicates how much response times vary from the average.

The table can be sorted using each column. Additionally, you can search within each column or within the entire table using the “search” box. The table can be exported to CSV, Excel, PDF. It can also be copied or printed using the links in the top left hand corner.




Holiday Biases


Lastly, the data was subsetted by date in an attempt to identify any patterns in holidays or time changes that may impact callback wait times. Calls are binned by calendar date and displayed as average callback time in minutes. All calls are categorized by calendar month.