The statistics Vibe probably has access to the widest range of information out of all of the Vibes. Within babblevoice, data is collected whenever a call hits a collection point. Depending on what happens in the process of handling the call, i.e. the call enters a queue or rings a phone, then different data is captured. This data is completely anonymized (excluding some data which might be phone centric for example).
Each report can show data in periods of an hour or day, then a graph can show a number of those periods, for example, a week’s worth. You can then look at an offset, for example - by the hour for 24 hours this time last week.
Reports can be visualised differently:
The last 2, Pie and Number, differ from the rest (well they are all different in their display) in what data they can display. With a bar chart, if you select an hourly statistic for 1 day, you will get 24 bars (or less if less data is available). With Pie or Number, it will only show the most recent data in the result.
Collection and the reporting of that data are broken down into 3 different sections:
Then for each section, there are multiple reports available.
We call them devices as they can represent servers as well as phones. Most users reading this document will be using them as phones, so for the purpose Vibes, consider a device a phone.
So, to get this straight in your head, imagine the phone on your desk. All the statistics based in this section are to do with the actions of using a phone - picking up a call, making a call and so on. On a device report You can create a report based on:
Each report has then sub items which are based on the first choice. So for example, most items contain a count by hour sub choice. So if you want to know how many calls were answered you have to choose Calls answered, then “count by hour”.
For the purposes of a phone, ringing could mean the phone itself ringing, or the phone has made a call and the phone it called is ringing.
With the devices report all the sections, except call spend, have the same sub reports, they are (each item can be viewed by hour or by day):
Bill seconds doesn’t specifically refer to time on the phone that you might receive a bill for. It means the time the call is connected - so historically may have received a bill for.
Duration of a phone call is from the start of dialling through to hanging up - so the ringing section and the talking section.
The number of calls in that period - so 10 calls in the hour.
The average length of time the phone is connected to another call. Measured in seconds.
The average length of time the phone is on a call, this is the sum of the ringing time and the connected time. Measured in seconds.
The sum of the ringing time for the phone.
The average amount of time the phone rang for the period.
Then we have call spend sub reports (again both types have options for a period for both hourly and daily):
The value of the most expensive call for the period.
The total call spend for the period.
Numbers are phone numbers you rent through babblevoice. Just the main phone numbers your customer calls. It can report on similar things to devices, but data is collected when similar actions are performed on your phone number (i.e. when someone calls it).
The following sub reports are available:
The total number of seconds of talk time for this number.
The total number of seconds of ringing time plus talk time for this number.
The call count for this period.
The average (mean) length of a call - for the talk time during this period.
The average (mean) duration - ringing plus talk time - of a call during this period.
The total sum of all ringing time of all calls for this period.
The mean of all ringing time of all calls for this period.
The sum of all ringing time for this period.
The mean of all calls for ringing for this period.
The highest cost of call for this period.
The total spend for this period.
Queues have 3 sub reports:
The sub-reports available for these 3 subtypes vary.
Currently unused and present for future development. A member is an on-hook agent ready to have calls dropped through to them.
Max number of consumers - phones ready to take calls for calls in the call queue.
The maximum number of callers in the queue during this period (waiting plus talking).
The maximum number of calls waiting in the queue during this period.
Bridges mean calls bridged to an agent. So agents having to talk on active calls. This is the maximum of bridged calls during that period.
The total number of calls that hit this queue for that period,
The lowest number of members during this period. Also currently unused - see the comments above.
The lowest number of consumers for this period.
The average number of callers for the period. This s the average of the number of callers in the queue taken every time a call enters this data collection.
The same as mean callers, but for waiting.
The same as mean callers, but for bridges - i.e. connected calls.
The maximum waiting time for all calls which gets answered from this queue.
The maximum talk time for all calls which gets answered from this queue.
The minimum wait time for all calls which gets answered from this queue.
The minimum talk time of calls which gets answered from this queue.
The average wait time of all the calls for the defined period.
The average talk time of all the calls for the defined period.
The total time waiting of all the calls which go through this queue then gets answered.
The total time of all the talk time of all calls which go through this queue then gets answered.
The number of calls which go through the queue which then get answered.
The maximum wait time for a call during the defined period.
The minimum wait time for the defined period.
The mean wait time during that period.
The total wait time across all calls for that period.
The number of abandoned calls for the defined period.
For all statistics we maintain 1000 data points. What does this mean?
If you are collecting the average wait time that a caller spends in a queue. Then we have to store the wait time for each call which waited in that queue. If you are interested in the wait time averaged over an hour, then this data is only meaningful if the number of callers is less than 1000 in that hour.
Once it has calculated the average this is then stored in another table. Again, we make space for 1000 data points for this transformed data. If we are looking at average per hour, then 1000 slots of 1 hour each, this means we have roughly 41 days of history.
The 41 days is the minimum you can rely on. If there are times where no phone calls happen, then that particular slot is not used in the table so another time gets to use that slow. For a business which only has office opening times then only data between Monday-Friday 09:00-5:30 will collect data, this greatly expands the history you will see in your Vibes report.