To explain the logic, we will use 3 segments, based on 2 Event Bands:

• A 13th June 2018
• B 13th June 2018
• D   13th June 2018

Each Segment is shown below. We will compare the 'Total Segments Less Dupes' number in each example.

Segment A

See screenshot below. 

There are 2 Include criteria set (see column 'Group' numbers 1 & 2). This structure is classed as an 'OR' statement - the customers attended an event in one event band OR attended an event in the other event band. This shows  a total segment number of 97,183.

Segment B

See screenshot below. 

The same 2 Include criteria are set as in Segment A, however, there is one difference - both criteria have the same number (see column 'Group' numbers 1 & 1), and are highlighted blue. This means they are grouped together. 

You achieve this grouping by selecting both criteria using the left-hand check boxes, and clicking 'Get People Who Fit All'.

This structure is classed as an 'AND' statement - the customer attended an event in one event band AND attended an event in the other event band. 

This shows a much lower total number of 37,376 which you would expect as the customer had to have attended an event in both Event Bands.

Segment D

See screenshot below. 

Segment D produces the same result as Segment A and is classed as an 'OR' statement - the customer attended an event in one event band OR attended an event in the other event band - however you'll notice that both Event Bands are in the one include criteria.

CASE 13525 

A client  hit upon an issue where he creates the two segments that theoretically should return the same count. In the screen shots above this would be B & C3  

Building a ‘criteria’ allows you to identify customers whose transaction / behaviour fit your criteria. 

Joining criteria together on the segment page only takes into account customers who exist in each criteria and pays no attention to those customer transactions  

Using Event Bands inside a criteria: 

When you select an event band inside a criteria e.g.  

You will be returned all the customers who made a transaction to an event in the event band.  The transactions are maintained so that when you click on to tab B “Customers Behaviour to selected Events” you can filter by the attributes of the transaction in tab A.  e.g. if you use the Visits filter on tab B to return customers who have made 2 visits – you would be returned customers who had made 2 visits to the events in the event band.

Using two event bands inside a criteria 

If you use two event bands inside a criteria like the example above it is looking for customers who 

have the same ‘TRANSACTION’ with in both event bands i.e. they have been to the same event in both event bands.  The Count number, however will only give you unique customers – but in the background only the transactions that match the criteria are used. 

Then when you filter by Visits in Tab B you can filter by the transactions that have been selected. 

Similarly, if you were to put both event bands into the same box (like below)  

You would for the customers who had bought for event in the event band only be able to monitor frequency and booking time in TAB B based on their transactions to those events. (i.e. if a customer had bought to events NOT in the event band these would not be included in the tab B filters.

Building multiple criteria lines. 

The crtieria lines do not consider transactions.  Therefore, if you have the following: 

The system is only looking at the customer records in each – here we are looking to create a unique list of customers.  That is if a customer exists in both criteria lines we should only count them once.  In this example if you simply add up the two criteria lines (72,271 + 59003 = 131,274).  However, when you click the refresh segment button – the segment count returns 93,898.  This shows that there must be people in both criteria lines e.g. 131271-93898 = 37376.  i.e. 37376 customers must exist in both crtieria lines.

By selecting the two tick boxes and clicking Get People Who Fit All – you will only have those 37376 customers who exist in both criteria lines.  i.e. the following screenshot returns 37376 customers