customer_features_daily
tablecustomerdaily
Daily booking and engagement features per customer. Anchor table for most ranking models.
warehouse.analytics.customer_features_daily
Trust
- Owner
- growth-team@example.com
- Lifecycle
- active
- Freshness SLA
- warn 36h · error 48h
- Last update
- 2026-04-26 20:38 UTC
Join keys
- Entity
customer_id- Grain
feature_date,customer_id- Timestamp
feature_date
Storage
- Row count
- 12,847
- Materialization
- table
- Source file
models/features/customer_features_daily.sql
Use this feature group
select
customer_id,
feature_date,
orders_count_7d,
orders_count_30d,
orders_count_legacy,
sessions_count_7d,
is_repeat_customer,
preferred_category
from warehouse.analytics.customer_features_dailySELECT *
FROM {{ ref('customer_features_daily') }}Features
| Name | Type | Column type | Null % | Distinct | Null behavior | Status | Declared ML consumers |
|---|---|---|---|---|---|---|---|
orders_count_7d
Number of orders placed in the trailing 7 days |
numeric | integer |
1.2% | 2,312 | zero | — |
churn_model_v2ltv_model_v3ranking_model_v4
|
orders_count_30d
v2Number of orders placed in the trailing 30 days |
numeric | integer |
1.2% | 2,312 | zero | — |
churn_model_v2ltv_model_v3
|
orders_count_legacy
Legacy lifetime count, kept for the now-retired ranking_model_v3 |
numeric | integer |
1.2% | 2,312 | zero | deprecated | — |
sessions_count_7d
Distinct web sessions in the trailing 7 days |
numeric | integer |
1.2% | 2,312 | zero | — |
churn_model_v2
|
is_repeat_customer
Has the customer ordered more than once |
boolean | boolean |
0% | 2 | — | — |
churn_model_v2
|
preferred_category
Most-ordered product category in the trailing 30 days |
categorical | varchar |
4.0% | 8 | propagate | — |
ranking_model_v4
|
Lineage
Upstream 2
-
stg_ordersmodel -
stg_sessionsmodel
Downstream dbt models 2
-
customer_features_lifetimemodel -
consumer_dashboard_metricsmodel
Auto-derived from the dbt graph.
ML consumers 3
-
churn_model_v2manual -
ltv_model_v3manual -
ranking_model_v4manual
Declared manually via used_by on individual features.