Overview
Segmentation in Corksy allows you to group customers based on operational, transactional, and behavioral data.
This article explains:
Where segmentation data comes from
The difference between static and dynamic segments
How and when data refreshes
What data fields are available
Clarifications based on common demo questions
Segmentation is designed to scale efficiently while abstracting underlying database complexity from end users.
Segmentation is currently in beta. The team is actively improving how segments are interpreted and populated, and an update to improve this behavior is in progress. If a segment's count looks lower than expected or a rule isn't populating the way you expect, that's the beta behavior — reach out to support and, if you need a reliable list in the meantime, use the Customers-page export workaround noted below.
Core Domains Supported
Customers
Memberships / Wine Club
Note: Migrated cancelled memberships do not sync
Orders (e-commerce, POS, manual, club)
Includes migrated orders
Order items
Order edits and refunds
Reservations
Customer behavior and engagement events
This supports both order-level and item-level segmentation.
When Is Data Written to ClickHouse?
Real-Time Sync
Segmentation data syncs from Corksy in near real time.
The following are synced automatically:
Orders (all channels)
Customers
Events
Club and membership updates
This ensures segmentation rules use the most recent operational data.
Initial Sync After Feature Enablement
When segmentation is enabled on an account:
A full initial sync occurs during the next scheduled sync cycle.
The scheduled sync job runs daily at 12:00 AM PST.
Daily Reconciliation Job
In addition to real-time syncing, a daily reconciliation job runs at:
12:00 AM PST
This job:
Verifies whether any records were missed
Automatically re-syncs failed or missing records
This ensures long-term data consistency and completeness.
Segmentation Types
Corksy supports two segmentation types, each with different behavior.
1. Static Segments
Static segments represent a fixed snapshot of customers at the time the segment is generated.
Behavior
Customers are captured once at creation.
The segment does not update automatically.
New qualifying customers are not added.
Customers who no longer qualify are not removed.
Best Used For
One-time campaigns
Historical exports
Fixed reporting snapshots
2. Dynamic Segments
Dynamic segments are continuously evaluated using the latest available data.
Behavior
Segment membership is recalculated using updated data.
Customers may be added or removed over time.
Segment counts may change based on new activity.
Dynamic Segment Refresh Behavior
When Is Data Evaluated?
For dynamic segments:
The segmentation list page does not fully evaluate rules.
Full rule evaluation happens when a segment is opened.
Refresh Button Behavior
The Refresh action updates only the segment count.
Opening the segment is not required to refresh the count.
Why Counts May Change
It is expected that:
The count shown before opening a segment may differ from the count shown after opening it.
This occurs because full rule evaluation runs when the segment is opened using the latest data.
This behavior is normal.
Segmentation Data Coverage
Segmentation operates on domain-level business data, not individual database tables.
Fields may be sourced from multiple ClickHouse datasets and are presented in a unified view.
Internal system tables are excluded.
Customer Data Fields
Customer segmentation supports:
Name, email, phone
Customer tags
Customer type (guest / registered / member)
Customer status
Lifetime value (LTV)
Total orders
Total spend
Last order date
Membership / Wine Club Data
Membership-based segmentation supports:
Club name
Membership name and level
Membership status (active, paused, cancelled)
Membership start and end dates
Customer–membership association
Membership shipping and billing addresses
Order Data (All Channels)
Order segmentation supports:
Order date and number
Order source (e-commerce, POS, manual, club)
Order total, subtotal, tax, shipping
Order frequency and recency
Refund and edit indicators
Order Item & Product Data
Item-level segmentation supports:
Product name / title
Product SKU
Quantity purchased
Item price
Total item value
Product-level purchase history
Customer Behavior & Engagement
Behavioral segmentation supports:
Cart abandonment status
Checkout activity
Page visit activity
Campaign engagement
Notification engagement
Reservation Data
Reservation-based segmentation supports:
Reservation date
Reservation status
Associated customer
Reservation activity history
External Platform Sync (Dynamic Segments Only)
Dynamic segments can sync with:
Klaviyo
Mailchimp
Sync Behavior
Sync runs once per day at 12:00 AM PST
New qualifying customers are added
Customers who no longer qualify are removed
Only dynamic segments are eligible for continuous sync
Static segments do not auto-sync after creation.
Need a reliable club-member list right now (e.g. for Klaviyo)? While segmentation is in beta, the most dependable way to pull your wine club members is to export directly from the Customers page: go to Customers, filter by Wine Club, and click Export. In the exported file, the "Accepts Marketing" column (column P) tells you which members are marketing-eligible.
Product Purchase Rule Clarification
"Customer Bought Product X"
When building segmentation rules based on product purchases:
If the product name is not provided exactly
And is not enclosed in quotation marks
Similar or partially matching product names may be included.
Best Practice
To ensure accurate results:
Use the exact product title
Enclose the product title in quotation marks
Example:
Customer bought product = "2022 Reserve Cabernet Sauvignon"
This prevents unintended matches when products have similar names.
"Bought one product but NOT another"
A common request is a list of customers who bought one specific wine but did not buy another specific wine (for example, bought the 2021 Cabernet Franc but not the 2022). Corksy does not currently support building that kind of "bought X and not Y" cross-product exclusion in a single segment or report. If you need this, contact support — it typically has to be assembled manually by exporting purchasers of each product and comparing the two lists.