Case study 03 / Segmentation
One list, four very different people.
Splitting a single blast list into behavior-based paths, including the one nobody builds for.
Segmentation path preview
- Sector
- B2B / PLG
- Lifecycle focus
- Across the lifecycle
- Read
- 8 minutes
The situation before the build.
This case study follows a product-led company whose audience has grown faster than its lifecycle strategy.
Most campaigns still treat the customer base as one broad list. But the people receiving those messages are no longer experiencing the product in the same way. Some have just started. Some are actively adopting. Some have stalled. Others are beginning to disengage.
One audience was hiding several customer journeys.
Sending one message to the entire database made campaign execution simple. It also meant the message had to be generic enough to apply to everyone.
A new customer might receive advice meant for an experienced user. A customer actively adopting the product might get a basic getting-started campaign. Someone who had stopped using a key feature could receive the same promotional message as the company’s most engaged customers.
The problem was not that the database needed dozens of segments. It needed a small number of meaningful customer states.
We organized customers around what should happen next.
The segmentation strategy started with lifecycle stage and recent customer behavior.
Rather than grouping customers by superficial characteristics, the system asks where the customer is in the journey, what they have done recently, and what the next useful action is for someone in this state.
Those questions create a manageable set of lifecycle paths: activation help for new customers, deeper value for adopting customers, recovery help for stalled customers, and retention for customers beginning to disengage.
Segmentation became part of the lifecycle system.
Eligible customers enter the segmentation workflow and are first evaluated by lifecycle stage.
Recent behavior then determines which path is most appropriate. Each path can change the message, the timing, or the next action while still remaining part of one overall lifecycle architecture.
The system also includes a none-met route. That branch matters because not every customer will fit perfectly into the conditions we expected when the workflow was built. Instead of disappearing from the logic, those customers become visible for review.
Segmentation lifecycle path
Data contract
What the workflow needs to know.
| Property | Mode | Source |
|---|---|---|
| lifecycle_stage | Reads | CRM property |
| last_meaningful_action_date | Reads | Product event |
| engagement_state | Writes | Segmentation workflow |
| segment_last_evaluated | Writes | Workflow timestamp |
Better segmentation should change customer behavior.
The test is whether behavior-based paths move more customers forward than the broad send they replaced.
We would hold a comparison group on the existing broad-send logic and measure downstream behavior, not just click rate. A useful segmentation system should improve the next customer action without multiplying sends or creating an unmanageable set of rules.
Customers reaching the next lifecycle milestone after entering a behavior-based path
Share of contacts classified into a deliberate path rather than the none-met route
Total message volume and unsubscribe rate must not increase simply because more branches exist
Your lifecycle
Have a similar problem in your system?
Tell us what is happening and what you want customers to do next.