Clean Room? More Like a Mud Room!

And you’ll need one soon

Neutral clean rooms, unique from their walled garden brethren, create horizontal integrations for marketers, publishers, and data providers to safely harmonize data across domains.

It is wholly understandable that the term clean room is used. The impending casualty of the third-party cookie has fomented disquietude in an industry with enough complexity and difficulty to make even the most assured somewhat anxious about any major change. Any opportunity to circumnavigate this negative outcome makes for a near spiritual moment.

“Our data is going to a clean environment to be washed away of all privacy violating sins.” Hallelujah!

With a more intricate understanding of what is happening within a neutral clean room, it is clear this description is not totally accurate. More precisely, it’s a place to leave your dirty (‘private’ in marketing parlance) data inside of an environment to be enriched and obscured, as to not destroy the rest of the house. Further, clean rooms can introduce data noise (more dirt) to further disrupt user level interpolation.

Sounds more like a mud room.

Closet Works defines a mud room as, “a place to remove coats and outer garments before entering the rest of the home.” Confirmed. The clean room branding is a misnomer.

So, What Is Happening in a Clean Room?

Unrelated interested parties push data into their own encrypted databases, where a shared attribute serves as a Rosetta Stone for computation, augmentation and deidentification. Strict access permissions are then implemented to provide appropriate utility of statistical knowledge (not raw data) to each party.

Not simple, but it’s the reason you need this technology. Let’s review:

The (Interrupted) Flow of Data

Whenever a web event takes place, data is generated. Carried through site code like tags and pixels, the information is stored in platforms after a user visits a website.

The data owner will attempt to use this data for marketing purposes. The data is collected by multiple code types, one of which is the third-party cookie.

The next step involves segmenting and packaging. The singular events are categorized so the cookies carrying specific attributes can become modeling seed data or lists for partners.

Naturally, data requires categorization for insights when shared. This sharing used to rely on cookie syncs; Mobile Ad ID (MAID) matching; and customer onboarding to connect disparate spaces, seamlessly. While imperfect, the process is well understood.

However, many of the controlling ecosystem players are now incentivized to restrict this flow of information. Identifier redaction from browsers and operating systems is impacting audience scale, conversion reporting, and attribution between tactics.

An Answer Emerges

Neutral technology allows for all the ecosystem players to combine this data with user privacy as a requirement.

The original data is now fortified with attributes. This takes the form of new users to target, or more knowledge about the current users in the file.

Finally, advertisers can further hide users’ identities, as the clean room technology adds extra user privacy to the original combined data set called differential techniques. These include, adding data noise; reducing trait precision; and including ID redaction thresholds.

The Saviors

Companies like InfoSum and Snowflake are making this a reality right now. Omnicom is engaged in real world scenarios where brands with sensitive data are currently using and testing these techniques to transact against, and augment, data that is currently a privacy quagmire.

How To Get Started

Every neutral clean room set up will be unique, but there are some general steps that one must always follow to get going.

Step One: Establish Client Objectives

The Analytics and Business owners need to work with Tech to establish objectives that are consistent with the Marketing Learning Agenda and available data variables.

Step Two: Set Up Client Account

Work with your Agency MarTech and Data Ops teams to enable the platform and set up the environment with the clean room provider, brand, and agency all aligned and properly provisioned.

Step Three: Data Integration

Tech will validate the match keys, encryption techniques, and data availability as Data Science begins to create queries to analyze the data against the agreed upon client objectives.

Step Four: Measure and Repeat

Use the data for the purpose of analysis of currently consented users and/or send new data to activation partners for communication.

And now you know what is coming very quickly, as deidentified data dries up. Just keep in mind, the clean room is more like a mud room since this is how the industry will maintain user privacy while amplifying marketer reach and knowledge.