How Media & Publishers Should Be Utilizing Snowflake Data Clean Rooms

Marketers rely heavily on attribution modeling to quantify the impact of each media campaign and measure their return on investment (ROI). This model uses third-party cookies to track people’s online activity and deliver personalized ads and content

However, these traditional methods of targeting and measurement have become scrutinized due to privacy concerns. Some of the key contenders in the advertising ecosystem have begun implementing consumer privacy practices to stay ahead of regulations. 

Take for example Apple requiring users to opt-in for their activity to be tracked across apps, and the forthcoming deprecation of third-party cookies on Chrome in 2023.  

With the deprecation of third-party cookies looming over the ad industry, marketers and publishers have shifted their focus to strengthening their direct consumer data relationships. A core pillar of that strategy is handling the data they collect directly from users and customers (i.e. first-party data) in a privacy-preserving manner. 

Data clean rooms have emerged as a strategy to do sophisticated audience segmentation, targeting, attribution, measurement, and machine learning-boosted lookalike analysis while preserving privacy. 

In this post, we’ll explain what a Snowflake data clean room is and how Media and Publishers can utilize it for improved advertising strategies.

What is a Snowflake Data Clean Room? Why Does it Matter to Media, Entertainment, and Publishers?

Snowflake data clean rooms are secure platforms that allow two or more parties to join and query data sets, without each party being able to see the other’s data. Snowflake data clean rooms allow for sensitive data derived from unique identifiers, such as emails, hashed emails, names, device IDs, and IP addresses, to be leveraged while preserving privacy. 

From there, marketers can segment and target existing customers by finding overlaps with a publisher’s audience without having to move or copy the data (i.e. ETL).

For a deeper dive into data clean rooms, be sure to check out our Introduction to Data Clean Rooms blog

How to Use Snowflake Data Clean Rooms in Media & Publishing

Step 1: Store Customer Data in Respective Snowflake Accounts

In this example, a retailer and publisher are launching a data collaboration. They can both keep customer data in their respective Snowflake accounts, with no need for further ETL. This makes it impossible for raw data to be viewed or copied by the counterparty without express permission.

Step 2: Join Data

The next step is to determine which can be joined and how to join them. They need to take into account any (personal) data restrictions and each of their goals. The simplest way is to join on data sets both already have, such as lowercase and trimmed email addresses or IP addresses. 

They can also use advanced forms of identity joining, graph joining, waterfall joining, and Boolean expression joining—or they can leverage a third-party identity provider on Snowflake Data Marketplace, such as Acxiom, LiveRamp, or Neustar.

Step 3: Calibrate

Next, they need to figure out how to query the joined data for the intended use case. It is important to define a query with the appropriate SELECT, GROUP BY, JOIN, and WHERE clauses.

For example, if a retailer wanted to query a publisher’s anonymized engagement data by content category and find overlaps with its own purchase data, the retailer would need to run a query and include the proper fields. The retailer can then use this information to better understand what types of content its customers consume.

Step 4: Approve

Afterwards, the requested query has to be approved by the counterparty, automatically ensuring that the query meets the counterparty’s rules for joint usage of its data. The approving party can set any rules to ensure the query meets privacy standards.

For example, one such rule could be to disallow any analyses or to suppress any output rows that aggregate fewer than 75 distinct people or devices in order to prevent unintended user re-identification.

Step 5: Run

The requesting party can now run the approved query across both its data and the other party’s data (as long as it respects the rules applied by the counterparty in the previous step).

Snowflake handles the computation of data between two (or more) parties.

Step 6: Activate

Aggregate-level results inform how audiences are segmented and targeted in upcoming campaigns. Take for instance the retailer in the previous example, they might learn that 7 percent of people who consume cooking content on the retailer’s site bought sneakers from the publisher in the past.

That segment can be targeted with ads for the latest models of sneakers in the future.

Step 7: Measure

After activation and the campaign has run (or while it is running), a party can perform another JOIN analysis to measure various campaign performance metrics.

Use Cases

Even though there may be some hesitancy when it comes to transitioning to newer technologies, rest assured that some of the largest customers in the industry are already utilizing Snowflake’s Media Data Cloud as the essential data platform for powering their advertising businesses, marketing execution, agency offerings, and data solutions. 

In 2021, NBCUniversal launched NBCU Audience Insights Hub, a platform built on a cross-cloud data clean room environment powered by Snowflake. In 2022, Omnicom Media Group became the First Agency Data Platform to Integrate with Audience Insights Hub. Omni has more privacy-compliant datasets than any competitive data offering, unlocking endless possibilities for strengthening direct consumer data relationships. 

Closing

It is only a matter of time before consumer privacy practices force media companies and publishers to re-evaluate the use of third-party cookies in their marketing strategies.

Leveraging Snowflake’s Media Data Cloud, any media company, publisher, advertiser, agency, or ad technology organization can design its own privacy and collaborative data environment for its own customizable solution.

Looking to make more data-driven decisions with Snowflake? phData can help! As the 2022 Snowflake Partner of the Year, phData thrives at helping media and publishing enterprises succeed with Snowflake.

Learn more today about our award-winning Snowflake consulting services!

More to explore

Accelerate and automate your data projects with the phData Toolkit

Data Coach is our premium analytics training program with one-on-one coaching from renowned experts.