counterweight/data-dwh-dbt-project
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merged PR 3666: Aggregated model for guest kpis metrics in intermediate

Browse source 
# Description

Aggregated model for guest kpis metrics.
This model aggregates the values for all basic metrics for guest kpis, it has a granularity per metrics of timeframe (YTD, MTD, WTD) and all other dimensions needed (has_payment, has_id_check & main_billing_country_iso_3_per_deal).
I decided not to include the calculated metrics because it adds much more complexity to the model specially when you start considering all possible combinations between dimensions and how this would affect the calculations, in this case conversion_rate, revenue_rate and average_csat_score. I think this could all be managed much simpler inside PBI and since the granularity of the model is not as deep as the daily model, this should help with the loading time for the report.

# Checklist

- [x] The edited models and dependants run properly with production data.
- [x] The edited models are sufficiently documented.
- [x] The edited models contain PK tests, and I've ran and passed them.
- [ ] I have checked for DRY opportunities with other models and docs.
- [ ] I've picked the right materialization for the affected models.

# Other

- [ ] Check if a full-refresh is required after this PR is merged.

Related work items: #24604
This commit is contained in:
Joaquin Ossa 2024-11-26 15:37:06 +00:00
commit dd4dff1479
2 changed files with 220 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

141
models/intermediate/kpis/int_kpis__product_guest_agg_metrics.sql Normal file
View file

@ -0,0 +1,141 @@
{% set yesterday = "(current_date - interval '1 day')" %}
{% set metric_names = (
"created_guest_journeys_not_cancelled",
"started_guest_journeys_not_cancelled",
"completed_guest_journeys_not_cancelled",
"created_guest_journeys",
"started_guest_journeys",
"completed_guest_journeys",
"total_csat_score_count",
"deposit_fees_in_gbp",
"waiver_payments_in_gbp",
"checkin_cover_fees_in_gbp",
"total_guest_payments_in_gbp",
) %}
with
int_kpis__product_guest_daily_metrics as (
select
date_day,
has_payment,
has_id_check,
main_billing_country_iso_3_per_deal,
{% for metric in metric_names %}
{{ metric }}{% if not loop.last %},{% endif %}
{% endfor %}
from {{ ref("int_kpis__product_guest_daily_metrics") }}
),
date_ranges as (
select
{{ yesterday }} as current_date,
date_trunc('month', {{ yesterday }}) as current_month_start,
date_trunc('week', {{ yesterday }}) as current_week_start, -- Start of the current week
date_trunc('week', {{ yesterday }} - interval '1 year') as py_week_start, -- Start of the same week last year
date_trunc('week', {{ yesterday }} - interval '1 year') + (
{{ yesterday }} - date_trunc('week', {{ yesterday }})
) as py_week_current_date,
date_trunc('week', {{ yesterday }} - interval '1 week') as pp_week_start, -- Start of the same week in the previous period
date_trunc('month', {{ yesterday }} - interval '1 year') as py_month_start,
date_trunc('year', {{ yesterday }}) as current_year_start,
date_trunc('year', {{ yesterday }} - interval '1 year') as py_year_start,
date_trunc('month', {{ yesterday }} - interval '1 month') as pp_month_start,
{{ yesterday }} - interval '1 year' as ppy_current_date,
{{ yesterday }} - interval '1 month' as ppm_current_date,
{{ yesterday }} - interval '1 week' as ppw_current_date,
extract(day from {{ yesterday }}) as current_day_of_month
),
aggregated_metrics as (
{% for metric in metric_names %}
select
'{{ metric }}' as metric,
has_payment,
has_id_check,
main_billing_country_iso_3_per_deal,
'YTD' as timeframe,
sum(
case
when date_day between dr.current_year_start and {{ yesterday }}
then dm.{{ metric }}
end
) as current_value,
sum(
case
when date_day between dr.py_year_start and dr.ppy_current_date
then dm.{{ metric }}
end
) as py_value,
sum(
case
when date_day between dr.py_year_start and dr.ppy_current_date
then dm.{{ metric }}
end
) as pp_value
from int_kpis__product_guest_daily_metrics dm, date_ranges dr
where extract(day from date_day) <= dr.current_day_of_month
group by
metric, has_payment, has_id_check, main_billing_country_iso_3_per_deal
union all
select
'{{ metric }}' as metric,
has_payment,
has_id_check,
main_billing_country_iso_3_per_deal,
'MTD' as timeframe,
sum(
case
when date_day between dr.current_month_start and {{ yesterday }}
then dm.{{ metric }}
end
) as current_value,
sum(
case
when date_day between dr.py_month_start and dr.ppy_current_date
then dm.{{ metric }}
end
) as py_value,
sum(
case
when date_day between dr.pp_month_start and dr.ppm_current_date
then dm.{{ metric }}
end
) as pp_value
from int_kpis__product_guest_daily_metrics dm, date_ranges dr
where extract(day from date_day) <= dr.current_day_of_month
group by
metric, has_payment, has_id_check, main_billing_country_iso_3_per_deal
union all
select
'{{ metric }}' as metric,
has_payment,
has_id_check,
main_billing_country_iso_3_per_deal,
'WTD' as timeframe,
sum(
case
when date_day between dr.current_week_start and {{ yesterday }}
then dm.{{ metric }}
end
) as current_value,
sum(
case
when date_day between dr.py_week_start and py_week_current_date
then dm.{{ metric }}
end
) as py_value,
sum(
case
when date_day between dr.pp_week_start and dr.ppw_current_date
then dm.{{ metric }}
end
) as pp_value
from int_kpis__product_guest_daily_metrics dm, date_ranges dr
group by
metric, has_payment, has_id_check, main_billing_country_iso_3_per_deal
{% if not loop.last %}
union all
{% endif %}
{% endfor %}
)
select *
from aggregated_metrics

79
models/intermediate/kpis/schema.yml
View file

@ -5304,6 +5304,85 @@ models:
Sum of total payments paid by guests, excluding taxes, in GBP
on the same date in the previous year, segmented by the specified dimension.
- name: int_kpis__product_guest_agg_metrics
description:
This model aggregates multiple metrics on a Year-to-date, Month-to-date or
Week-to-date basis. This model changes the display format of the model
int_kpis__product_guest_daily_metrics pivoting the metrics columns and
adding a timeframe dimension.
columns:
- name: metric
data_type: text
description: Name of the business metric
- name: has_payment
data_type: string
description: Has there been any guest payments on the guest journey.
tests:
- not_null
- accepted_values:
values:
- W/O Payment
- With Payment
- name: has_id_check
data_type: string
description: Does the verification in the guest journey
includes Government Id Check for the bookings.
tests:
- not_null
- accepted_values:
values:
- W/O Id Check
- With Id Check
- name: main_billing_country_iso_3_per_deal
data_type: string
description: |
Main billing country of the host aggregated at Deal level.
tests:
- not_null
- name: timeframe
data_type: text
description: |
Timeframe considered for the aggregation, it could be Year-to-date,
Month-to-date or Week-to-date
tests:
- not_null
- accepted_values:
values:
- YTD
- MTD
- WTD
- name: current_value
data_type: numeric
description: |
Numeric value (integer or decimal) that corresponds to the timeframe
computation of the metric at the current date.
For example if the current date is 27/11/2024 and the timeframe is MTD,
then this value would correspond to the computation of the metric for
the dates between 01/11/2024 and 27/11/2024.
- name: py_value
data_type: numeric
description: |
Numeric value (integer or decimal) that corresponds to the timeframe
computation of the metric at the current date but on the previous year.
For example if the current date is 27/11/2024 and the timeframe is MTD,
then this value would correspond to the computation of the metric for
the dates between 01/11/2023 and 27/11/2023.
- name: pp_value
data_type: numeric
description: |
Numeric value (integer or decimal) that corresponds to the timeframe
computation of the metric at the current date but on the previous period.
For example if the current date is 27/11/2024 and the timeframe is MTD,
then this value would correspond to the computation of the metric for
the dates between 01/10/2024 and 27/10/2024.
- name: int_kpis__metric_daily_new_dash_created_services
description: |
This model computes the Daily Created Services at the deepest granularity.