How to add custom Programmatic cost in ADL

Overview

ADL 2.0 supports the most popular advertising platforms including the biggest DSPs used in programmatic advertising. However, not all DSPs currently have a datasource connect within Alli Data. In these cases, customer teams often build custom e-mail or SFTP datasources to ingest data from various programmatic vendors/publishers.

Learn how to add this data to the ad_performance and creative_performance report with just a few steps.

Prerequisites

ADL reports exist for your customer (see Getting started with ADL)
Understand how to create a view in Alli (see next section)

How to create a view…

in Redshift (using Custom SQL option)

Getting started with Explorer Reports

in BigQuery

Create BigQuery View

Instructions

Before proceeding, ensure that view or datasource doesn’t already exist for your customer. You can do this by querying select count(*) from {alli_client}.programmatic_cost in the data warehouse environment of your choice. If it already exists, you can modify the view as necessary.

1.) Create a datasource or view named programmatic_cost within the data warehouse environment of your choice with the following schema:

Field name	Data type	Description
`date`	DATE	The spend date Ensure that this matches the customer’s CM360 account timezone (typically set as `America/New_York` / `EST`)
`platform`	STRING	The name of the DSP, vendor, or publisher
`placement_id`	STRING	The associated CM360 Placement ID
`cost`	FLOAT	The spend amount
⬇️ (Optional) ⬇️
`impressions`	INTEGER	The number of impressions
`clicks`	INTEGER	The number of clicks

2.) Validate that the table or view exists in the data warehouse:

Checking that the data exists in Redshift in Data Explorer

3.) Run your customer’s ADL 2.0 - Refresh app with DEPLOYMENT set to True

4.) Validate that the cost associated with the corresponding platform is properly added to ad_performance as expected, e.g.:

SQL

select
  platform,
  source,
  sum(cost) as cost
from playground_core.ad_performance
where source like '%programmatic_cost%'
and platform = 'AdTheorent'
group by 1, 2;