Streaming Spend and Impressions Metrics

The current version below is v2. To access the previous version, click here.

Tatari offers granular TV response data via Tatari S3 buckets. S3 is a secure cloud file-sharing service that allows you to both download raw data and automate ingestion (similar to an API) into your multi-touch attribution or internal tools. Learn about streaming spend and impression metrics below.

This article contains the following:

Overview of spend and impressions data
Overview of the files
Where to find the reports
Getting started: How to access the reports

Overview of spend and impressions data

Spend and impressions data allows you to see how much an ad cost, how many people saw it, and how it performed against other spots.

Overview of the files

You’ll have access to two streaming spend and impression files via Tatari S3 buckets:

Daily file (immature): This file shows immature data—or the data that we measure before receiving the final spend costs from the publisher— from the past 45 days*. For example, if the file was created on June 1, 2020, it will contain data from April 17 - June 1, 2020 (the past 45 days). The daily files may have missing or sparse data as we wait for actuals from the publisher, but will allow you to access immature data until the matured data is available from the monthly files.
Monthly file (mature): This file shows mature data for the previous month*. For example, final spend and impressions data for the month of January (the previous month) would be available March 1st.

*All files are inclusive of data on the start date and exclusive of data on the end date.

File example

Within the CSV file, you’ll find the following data broken out by creative:

Column	Type	Description
date	date	Date when the streaming occurred* (UTC)
platform	string	Where the ad was aired
measured_spend	double	Spend as per original agreement (CPM x impressions/1000)
booked_cpm	double	Cost per 1000 impressions (i.e. the amount of spend to secure 1000 impressions); agreed to ahead of time
impressions	integer	How many times the ad was displayed
effective_spend	double	Spend calculated using effective_cpm when mature spend data is available. Otherwise, this calculation will utilize booked_cpm to calculate spend
effective_cpm	double	Actual CPM; this comes once we receive the platform invoices and reconcile the streaming spend; it could be adjusted if there are bonus impressions
creative_code	string	Creative ISCI (Industry-Standard Coding Identification for the commercial)
campaign_id	string	Streaming campaign identifier; this can be used to link to data in other files
creative_name	string	The name of the creative

*If the date here shows 4/12/20 for example, then the streaming occurred during that day anytime from 4/12/20 at the 0:00 hour to 4/12/20 at the 23:59 hour eastern time.

How can I calculate CPV & CPA metrics?

To calculate the CPV and CPA, you can match on the campaign_id in the streaming spend and impressions files available in S3 buckets.

Example streaming spend and impressions file

How often files are generated

The files are generated as follows:

Daily file (contains 45 days of data): Every day from 20:00 - 22:00 UTC
Monthly file (contains 1 month of data): The first day of the month from 20:00 - 22:00 UTC

Some examples:

Daily file: If it was created on June 1, 2020, it will contain data from April 17 - June 1, 2020.
Monthly file: Final spend and impressions data for the month of January would be available March 1st.

All files are inclusive of data on the start date and exclusive of data on the end date.

The measurement attribution type

The streaming spend and impressions report is methodology agnostic.

How to keep reporting up-to-date

To see the most recent data in your internal reporting, you should overwrite the previous days' files. We recommend you replace all data from the date ranges in the daily file (45 days immature data) every day once you receive the new file and then again once you receive the monthly (1 month mature data) file.

You can also refer to our sample python script here, which pulls the report data from our S3 bucket and creates an SQLite3 database locally that you can query. In the script you will find functions that show you how to properly overwrite existing data for initial and daily incremental loads, and thus would be a good reference when building out your own automated process.

Where to find the reports

You can find streaming spend and impressions reports in the Tatari S3 bucket under the folder streaming_spend_and_impression/v2.

Getting Started: How to access the reports

Follow the instructions in your dashboard’s AWS Data Sharing page to gain access to your reporting.

Log in to your dashboard
On the bottom left navigation bar, select “Settings” -> “Integrations”
Select "AWS Data Sharing" from the top navigation bar.