Data Profiling - Grouped Outcome

1 Overview

Outcomes in bxp software (bxp) are very powerful. Extending their power even further into their reporting capability is a great addition to their capability set.

Outcomes previously were grouped as Complete and Incomplete. This limited reporting capability. It is now possible to put two separate levels of customisation for grouping in automatic reports directly into the outcomes themselves.

2 Understanding what grouping is about

In order to realise the true value of where grouping lies we need to understand what grouping is about. For this we can use a worked example. We have a sales campaign with the following outcomes:

• Contact - Callback - Afternoon
• Contact - Callback - Evening
• Contact - Callback - Morning
• Contact - Callback - Personal
• Contact - Emergency escalation
• Contact - Not interested
• Contact - Sale
• No Contact - Disconnected tone
• No Contact - Flagged exception
• No Contact - No answer
• No Contact - No phone number
• No Contact - Voicemail
• No Contact - Wrong number

Work is done in this form and we now have a few hundred records all with different outcomes. As a manager we're only interested where there was a legitimate opportunity to sell. We need a report to show, that when there was an opportunity to sell, how well we did. Previously this would require a custom report to be built. Instead, of the outcomes there we ask "Which outcomes do constitute a genuine opportunity to sell?" We come upon the following:

• Contact - Not interested
• Contact - Sale

So we can group these two outcomes with "Opportunity to sell" and all the other outcomes with "No opportunity to sell"

The data quality manager come along, but wants a different perspective of the same data. How much of the data was non-contactable? So using a second grouping we identify the following outcomes as bad data.

• No Contact - Disconnected tone
• No Contact - Flagged exception
• No Contact - No phone number
• No Contact - Voicemail
• No Contact - Wrong number

As can be seen, not all of the No contact outcomes are included. So with these outcomes we can flag them as "Bad Data outcome" vs. "Good Data outcome".

We have different perspectives on the same data. bxp can handle two separate levels of grouping on the same outcomes. The report engine in bxp can look at the wording applied to the groups and generate the reports automatically with drill down into the groups.

3 Setup

3.1 Choosing your groupings

The first step is to group your outcomes. This is a matter of choosing the words which will best describe the grouping for your reporting. Some examples are:

For Sales, this grouping allows drill down on just the outcomes where the agent has the opportunity to perform a sale:

• Opportunity
• No Opportunity

For Customer Care, this grouping allows focus to be put on actual outcomes where the contact is meant for the care department and not other departments

• Care Contact
• Non-Care Contact

For Data Quality, this grouping allows focus to how much waste in the records there was, which can be especially important for outsource companies trying to hit conversion targets

• Bad Data
• Good Data

There are many options and the outcomes are not limit to only two.

3.2 Update the outcomes

Main Menu > Form Management > Form - Outcome Manager -> Choose the form > Choose an outcome > In the field 'Report Grouping 1' add the Grouping name appropriate as decided above

You need to repeat this exercise for every outcome

These groupings can be changed at any time, in real time and the reports will update accordingly.

4 Using the reports

The ground work done, the reports can now be generated.

Main Menu > Data Profiling > Instant - Statistics > Statistics - Grouped Outcome - Report by Period > Choose the period, Choose the form and Choose "Grouping 1"

In this example the groups are

• Opportunity to Sell are outcomes where there was an ability to chat to someone who could buy
• Worked Lead are outcomes where there was a no opportunity to sell
• System are system generated outcomes such as "Taken to process", Clickthrus and other automated processes
• Others are non-grouped outcomes

Hovering the mouse of sections of pie provides details on that slice of the pie.

With the groups on the left there are drill down links to allow you to see how that group was made up.