A Data Clean Room is a secure, protected environment where PII (Personally Identifiable Information) data is anonymized, processed and stored to be made available for measurement, or data transformations in a privacy-compliant way. The raw PII, is made available to the brand and is only viewable by the brand.

How does a Data Clean Room work?

All user-level first-party data loaded from CRM systems (including historical data) like Salesforce, or ecommerce platforms (such as Shopify, Magento, Epsilon). are loaded into this secure environment. Any other data sources including historical and current transaction data can also made available in the clean room environment for a variety of use cases.

The PII data sent to the clean room is hashed for transmission and once it enters the clean room it is secured and encrypted, protecting it from unauthorized access. Brands have full control over the clean room, while partners can get a feed with hashed PII data as an output. This anonymized data can then be shared in a compliant way with measurement partners like Measured or media/publisher platforms like Facebook and Google.

What’s the benefit of a Data Clean Room?

It is set up by a partner like Measured, but is handed over to the brand to use as a turnkey feature giving complete control of the environment to the brand.

What are some additional privacy features of a Data Clean Room?

A consent management system that captures first-party acceptance of cookies to be added to the environment which assists with adherence to CCPA/ CPRA /GDPR regulations. These consent signals are applied to data in the Clean Room, resulting in enriched data that can be used for measurement, or to pass back to media/publisher platforms.

All in all, the Data Clean Room is a trusted Turnkey Compliance solution that is easy to deploy and maintain within your secure environment for privacy compliance.

Use cases for a Data Clean Room:

• Anonymizing user-level PII data that can be used for measurement
• Automation for upload of offline data to publishers like Facebook for matchback processes
• User-level analysis of customers including LTV reporting
• Cohort level analysis.
• Built-in privacy compliance support for regulations such as CCPA, CPRA and GDPR.

This can be done in a number of ways, but there is no doubt that best-in-class cross-channel reporting is created by collecting and aggregating data via direct vendor-level API based data feeds updated at the daily level.  For vendor without API based data feeds automated email to FTP uploads minimize latency and maximize data accuracy compared with manual data ETL operations.

Marketing ROI (Return On Investment) is calculated as the ratio of incremental contribution to sales from a marketing channel divided by the spend in that marketing channel. This is the most common formula used to evaluate the return on investment of media dollars.

Incremental_sales = f(media_spend, raw_sales, other_parameters)

Once incremental sales is calculated for a tactic using one of the above techniques, the ROI for that tactic is calculated using the following formula.

ROI = (incremental_sales) / (media_spend)

Or

ROI = (revenue – media_spend) / (media_spend)

Another commonly used marketing term is ROAS: Return On Ad Spend. The formula for ROAS: (revenue) / (media_spend)

Yes, incrementality can be measured on Direct Mail and Catalog. It is in fact among the more measurable channels because of the strength of identity resolution between audiences who received a mailer and customers transacting online, or in store.

The classic approach to incrementality measurement on Direct Mail and Catalog is to use systematic holdouts and compare the response rate and revenue per piece from users in the holdout group versus the users in the mailed group. This will get you the true measure of Direct Mail and Catalog impact to the business. You can read more on incrementality testing here.

The most basic form of performance reporting on Direct Mail and Catalog is through what is called matchback reporting. The list of households that were mailed a piece is matched to the list of customers who transacted, to identify how many of those in the mailed cohort eventually made a purchase. Brands who are more analytical holdout a cohort of users from the households selected to be mailed. Metrics from the matchback reporting on mailed users is compared to the same metrics on held-out users to calculate the incremental impact of the Direct mail or Catalog campaign.

There are two major types of Direct mail and Catalog campaigns.

1. Housefile campaigns are programs that target households who are typically customers of the brand, and the brand has both the household address and consent from the user to market to them.
2. Rental campaigns are programs that target households who are typically prospects of the brand. The names and households targeted are typically rented from a co-op that has the household address and has obtained consent from the user to be marketed to.

The audience selection, response rates, mail merge process, data collection process, data formats, data quality, match back process and other associated norms vary significantly between Housefile and Rental campaigns. Hence, experimental design, holdout selection and data processing is tailored very carefully to align with the process behind how lists are created and households are mailed for housefile and rental campaigns.

For Direct Mail and Catalog, the incrementality reads vary significantly by several audience dimensions – Housefile vs Rental, recency of engagement with brand, frequency of engagement with brand etc., Also, it is reasonably typical to find that response rates and incrementality are inversely correlated.

There are many challenges to measuring marketing effectiveness. Factors include:

• Types of Data Available: The type, richness and quality of data vary widely from channel to channel. For example, with programmatic display advertising, marketers can get user-level and impression-level data to track the user all the way to a conversion, regardless of whether a conversion occurs or not. However on TV, marketers would only get reporting of when their ads were flighted on air. On Facebook, marketers collect rich aggregate data reporting about their targeted audience, but can obtain almost no user-level data.
• Frequency of Data: Some media channels offer data on a real-time basis, whereas others only offer data post-campaign and reporting may only be delivered weekly or monthly.
• Attribution and Incrementality Methods: Not all attribution models and incrementality studies are created equal. Depending on the type of methodologies used, the channels under test and a host of other factors can lead to limited inputs for a media measurement framework. Gauging the contributions of each media channel, tactic, or audience without an always-on test and control experimentation practice in place is very difficult to execute in an apples to apples way.

For years marketers have been searching for the holy grail of media measurement, one that could address the challenges stated above. A single framework that ingests all online & offline data and automatically generates reports and insights to guide budget reallocation decisions in real-time with no friction. At one time we thought MTA was the answer to this desired state, however it failed to fulfill its promise due to many factors like – the collapse of the third-party cookie, identity resolution gaps, severe data reconciliation issues and perhaps the biggest blindspot of all, no access to the walled gardens. For all intents and purposes MTA is no longer a viable exercise.

The types, frequency and availability of user-level data limits the universe of advanced measurement techniques that can be used to measure marketing effectiveness. The types of decisions that the measurement informs also plays a big role in how useful marketing measurement itself is. Marketers looking to make tactical decisions, like optimizing between creative A vs creative B running on a specific audience within a specific media channel, can use data made available by that channel.

There are advanced measurement methods available that can answer the cross-channel attribution problem statement but are not subject to the limitations of MTA. For more strategic decisions, like budget allocation across multiple media tactics based on incremental sales and incremental ROI, you would have to use advanced marketing measurement techniques such as incrementality measurement to get to an answer. For other big strategic questions, like the impact of weather, competition, interest rates, or government policy on sales would need to apply a marketing measurement technique like MMM (Marketing Mix Modeling) for long-term planning.

Here are additional resources for reference:

• Incrementality Testing
• Marketing Mix Modeling (MMM)
• Multi Touch Attribution (MTA)
• Pros and Cons of Incrementality Testing vs MMM vs MTA