By using Mautic, I’ve been faced with something that I’ve never thought about: segmentation based on behavior.
I can imagined (and, I fact I needed) a few useful use cases:
Contacts that visited landing pages X in the last Y days
Contacts that had filled form Z
Contacts that visited URL X
It would be a huge advancement if we could filter contacts based on behavior. A good starting point would be the possibility of adding filters based on contact’s action on Mautic’s components.
Just to share with you, one good reference to look at is SAP hybris Marketing, which segmentation engine is certainly one of the best when compared to the comparable high-end solutions.
By using Mautic, I’ve been faced with something that I’ve never thought about: segmentation based on behavior.
I can imagined (and, I fact I needed) a few useful use cases:
Contacts that visited landing pages X in the last Y days
Contacts that had filled form Z
Contacts that visited URL X
It would be a huge advancement if we could filter contacts based on behavior. A good starting point would be the possibility of adding filters based on contact’s action on Mautic’s components.
Just to share with you, one good reference to look at is SAP hybris Marketing, which segmentation engine is certainly one of the best when compared to the comparable high-end solutions.
@paulofelix - Agreed. This would be a great feature for Mautic, and is an essential feature for nudge based Marketing Automation. All kinds of on-page events could be triggered via plugins based on these behavior segments. Content Recommendations, “Resume where you left off,” and more targeted messaging or offers are just a few options I can think of off the top of my head.
If the Mautic Dev Team brought back PostgreSQL support, it would also be possible to apply machine learning based segmentation in Mautic using Apache MadLib. This would be especially useful for deploying Content Recommendations using Behavior #3. Even without PostgreSQL, it would be possible to tap the DB and run external machine learning using these segments.
In fact, @ninjoan suggestion is a little bit different… You should use a “general segment”, then, while creating de decision tree into the campaign, insert a decision based on Visit Page… This combination of a broad segment + the campaing decision is the “fake” behavioral segmentation…
It seems to work - however, as I said, is just a workaround… Rgrds
I’m fairly new to Mautic but have been using ESPs and enterprise MA tools (Adobe, Oracle, Salesforce, IBM) for the last 5 years. Mautic has some impressive features and in places goes head to head with the likes of Oracle and Salesforce offerings. The one feature that seems to be missing is the ability to create custom data objects and then use that data to apply rules in segments, dynamic content and campaigns. For example I may want to reference a customers purchase behaviour or a a product recommendations list or promotions data in segmentation which as you mention may have been generated using machine learning outside of the tool.
Appreciate there may be workarounds as described and contact fields can be added, data can be preposed outside and populated in the contact object but this is cumbersome - having the option to create custom objects and reference 1 to many relationships to the contact is useful for a number of use cases.