Choosing the proper segmentation rule

After creating a segmentation, it will be necessary to create an automatic segmentation rule if you want it to be dynamic and calculated once per day.

At the second step of creating this rule ("Segmentation"), you will have to define the action that you want to apply to profiles. Three kinds of action are available:

Add: Profiles who meet the targeting criteria are added to the segment
Remove: Profiles who meet the targeting criteria are removed from the segment
Match: Profiles ending up in the segment will be exactly those matching the targeting criteria. This implies both adding profiles who meet the targeting criteria and removing profiles who no longer meet these criteria.

images/download/attachments/615294114/image2022-6-22_10-22-24.png

It can be tricky to instinctively choose the most appropriate of these actions. Indeed, the segmentation type often determines which action is the most relevant.

Simple segmentation

A simple segmentation groups profiles sharing common characteristics in a segmentation containing a single segment. The affiliation of a profile is characterized by true or false. As such, defining this kind of segmentation is usually fairly straightforward: either a profile belongs to it, either he does not.

Recommend action: Match

Due to the nature of simple segmentations, the most appropriate action is usually to choose "Match". Indeed, either the profile belongs to the segmentation, either he does not. In this case, it is rarely necessary to keep a profile in the segmentation if it does meet the defining criteria anymore.

A profile matching the targeting will be therefore added to the segmentation, while a profile who does not match this targeting anymore will be removed from this segmentation immediately and without ambiguity. As there is in fact only one segment in this segmentation, choosing where to transfer removed profiles does not come into question: they are simply not in the segmentation anymore.

Specific case: Add

The "Add" option may prove to be relevant for simple segmentations when it revolves around a historical aspect of profile behavior. In such a case, this segmentation is unchangeable and profiles are not meant to exit this segmentation. It allows to target profiles who once made an action or who once had a specific behavior in the past, independently from their current behavior.

For example:

"Profiles who have made a first purchase": The total amount of purchases of a profile does not matter, only the fact that they made a first one is relevant for this case.
"Profiles who have owned a vehicle from the Blue brand": It does not matter that a profiles currently drives a car from the Red brand, as in the past they once chose a Blue car.

In the case of a segmentation based on history, we suggest you to choose the "Add" action rather than "Match". There are two main reasons for this:

On the one hand, removing profiles is not useful for this kind of segmentation
On the other hand, there is a risk with 'Match' of accidentally removing the profile should the information source not appear anymore in ACTITO (for example, if entries of a custom table are archived).

Specific case: Remove

The "Remove" action is a special case in the sense that it will be only used in specific contexts, as completely removing a profile from a segmentation is rarely repeated as an automatic process.

"Remove" can be used in two specific cases:

Correcting your segmentation

This use allows you to correct the affiliation to your segmentation for profiles who were added to it while they should not have been.

In such a case, it will be a one-off rule rather than an automatic rule. You will need to proceed exactly as for creating an automatic segmentation rule, except that you will not check the "Activate automatic segmentation" box at step 4.

Automatic rule with "Member by default"

There is a specific case for which using an automatic segmentation rule with the "Remove" is appropriate: when you created a simple segmentation for which every new profile is "member by default". To do so, you need to check the "Member by default" box while creating a simple segmentation. This setting is useful for communication aimed at new profile, especially if you are waiting for an action on their end or from your team.

In such a case, you can create an automatic segmentation rule to remove the profile from this segmentation as soon as it meets the qualification criteria specific to your activity.

For instance:

"Profiles created during the previous month": This kind of segmentation lets you target recent profiles. You can create an automatic rule to remove them from the segmentation once their creation date is more than 30 days ago.
"New subscribers without purchases": This kind of segmentation lets you target profiles who registered to your e-commerce website without purchasing anything yet. You can create an automatic rule to remove them from the segmentation as soon as any purchase is recorded in a custom table.
"Profiles without sales representative": This kind of segmentation lets you target profile without any dedicated contact person yet. You can create an automatic rule to remove them from the segmentation when a representative is assigned to the profile via a linked data table.