Addressed comments

models/intermediate/kpis/int_kpis__dimension_date_product_guest.sql

@@ -1,5 +1,3 @@
-{% set starting_date = "('2020-01-01')" %}
-
 with
     unique_dimensions as (
         select distinct has_payment, has_id_check
@@ -8,4 +6,4 @@ with
 select d.date_day, ud.has_payment, ud.has_id_check
 from {{ ref("int_dates") }} as d
 cross join unique_dimensions as ud
-where d.date_day >= {{ starting_date }}
+where d.date_day >= {{ var("start_date") }}

models/intermediate/kpis/int_kpis__product_guest_daily_metrics.sql

@@ -72,6 +72,7 @@ with
 
 select
     gk.date_day,
+    gk_py.date_day as py_date_day,
     gk.has_payment,
     gk.has_id_check,
     sum(
@@ -130,4 +131,4 @@ left join
     and gk.has_id_check = gk_py.has_id_check
 -- Here we use a group by because of leap days that generate duplicity when being
 -- moved to another year
-group by gk.date_day, gk.has_payment, gk.has_id_check
+group by gk.date_day, gk_py.date_day, gk.has_payment, gk.has_id_check

models/intermediate/kpis/schema.yml

@@ -5830,6 +5830,7 @@ models:
       The unique key corresponds to the deepest granularity of the model,
       in this case: 
       - date_day,
+      - py_date_day,
       - id_deal,
       - has_id_check.
 
@@ -5837,6 +5838,7 @@ models:
       - dbt_utils.unique_combination_of_columns:
           combination_of_columns:
             - date_day
+            - py_date_day
             - has_payment
             - has_id_check
     columns:
@@ -5846,6 +5848,12 @@ models:
         tests:
           - not_null
 
+      - name: py_date_day
+        data_type: date
+        description: |
+          Date on the previous year of when Guest Journeys have been completed.
+          Note that this date can be NULL for leap days (29th February)
+
       - name: has_payment
         data_type: string
         description: Has there been any guest payments on the guest journey.