What data clean rooms mean for the privacy-first internet
Data clean rooms enable marketers to access scores of valuable consumer data in order to optimize their targeting efforts — in an anonomized and secure way that speaks to growing privacy concerns. But marketers have many different types of data clean rooms to choose from and the best-fit will vary depending on a range of factors, writes Katherine Strieder, global chief product officer at programmatic marketing firm MiQ.
The digital ecosystem is currently witnessing a mass expansion of consumer experience as well as a speedy evolution of privacy and regulatory compliance. As someone who once activated audiences by uploading lists of content categories into platforms serving display banners (long before privacy and data-sharing laws), I often think about how we got here.
Today, the walled gardens of Facebook, Apple, Amazon own large swaths of customer data, content as well as much of the core technology, all of which drives the industry. Smart companies now realize the value of their own data, and technology vendors are evolving as audience data is applied in new and innovative ways.
The latest marketing news and insights straight to your inbox.
Get the best of The Drum by choosing from a series of great email briefings, whether that’s daily news, weekly recaps or deep dives into media or creativity.Sign up
Meanwhile, the industry is evolving its business and ethical standards to mirror those which have long governed sectors like healthcare, traditional TV and so on. These structures weren’t originally embedded in the digital ecosystem for a variety of reasons. Some posit that publishers and brands didn’t set up enough governance with the tech stacks. Others suggest that the leadership building today’s walled gardens were opportunistic without considering the conflicts of interest associated with powering tech, content and consumer data. What’s more, the space evolved much faster than regulators usually move.
Because consumers operate primarily in the digital universe today, their data is the most effective way by which companies can reach them. Methods of doing so have expanded as mediums of engagement have multiplied and privacy practices across markets and verticals have diversified. The question for marketers then is which data and methods to use.
Enter clean rooms.
Of many methods, one is the data clean room. Using nascent technologies, it provides a unique opportunity for marketers to connect different data sets, graphs, panels and experimental data science to build insights about consumers on a global scale — in privacy-first, secure environments. However there are many types of clean rooms, and it’s important to understand their differences.
Walled garden data clean rooms
Walled gardens have an interest in maintaining their dominion over tech, content and data. However, they also have an interest in engaging other companies to reach consumers. On the flip side, companies need access to consumer data made available via these walled gardens. As such, tech players have made forays in data clean rooms, which allow marketers to analyze their own first-party data against walled garden data, without either side actually handing that data over.
When partnering with these vendors, it’s key for marketers to protect their data and insights in very clear terms. They must also understand that the primary goal of these clean rooms is to drive more buys on their platforms.
Google’s Ads Data Hub (ADH) is a good example of one such clean room. Within this space, a clothing brand could port its own purchase data into Google ADH and match it with Google campaign data to find patterns between time of purchase and creative served. The brand may find that, contrary to its original assumption, a jeans and t-shirt creative combo converts just as well on weekdays as weekends. It can then alter its campaign accordingly.
Amazon has also recently launched its own version. In the coming months, however, we’re likely to see companies without the tech dominance — but with the critical mass of users and content (especially those with valuable streaming content like Disney, Spotify and TikTok) — start building their own clean rooms.
Neutral data clean rooms
Neutral data clean rooms have also entered the space. They work across markets, verticals and mediums. They allow for virtual data collaboration using multi-party encryption technology and data science. Examples of neutral data clean rooms include Nth Party, Kochava, Infosum and Snowflake. These platforms operate very differently than the clean rooms governed by major tech companies and operate under their own unique business models.
These clean rooms all experiment with big data to surface innovative data insights. Companies can securely and privately connect, analyze and enrich their data with datasets, graphs and integrations from other sources — without exposing or moving the underlying. This flexibility enables a variety of use cases for marketers to understand consumers in the digital universe.
For example, a car rental company can target interested customers arriving at JFK Airport in New York. The airlines operating out of JFK know who is flying in. By combining anonymized car rental data with airline data, clean rooms cross-reference a list of loyal rental customers with a group of travelers coming to JFK for targeting. This approach benefits the airlines, too. The data from the rental car company identifies anonymized groups that arrive via other airlines as well. As such, airlines can build strategies to target these groups.
So, what should your clean room strategy be?
All digital marketers need to consider how to fit clean rooms into their vendor ecosystem to evolve their audience strategies.
Some useful questions to ask when thinking about your strategy include:
- How advanced is the vendor in flexible metadata layers? Does it support multi-client data ownership? If not, it is more of an identity resolution engine suitable for a single business with diffuse data assets?
- Does the vendor enable choosing and fusing data and graphs to associate data between parties?
- Which data sets and graphs does the vendor work with and in which markets?
- How flexible is the vendor in allowing data science with raw data?
- Does the vendor offer a self-service user interface and a data science environment to manipulate or run models?
- How well-versed is the vendor in privacy across different markets and verticals? All clean rooms and data bunkers offer varying levels of data security.
- This space will continue to evolve. Partners will change over time. Can you use your influence to move these companies to better business decisions for the industry?
- As with any digital marketing technology, the best-fit solution for a given marketer and the companies with whom the organization collaborates will depend on the overall business strategy and desired outcomes.
What is certain is that marketers will need a strategy. The high expectations around customer experience coupled with the rapid evolution of privacy-first advertising means that the old ways of doing things just won’t cut it. Marketers need to be figuring out how to make the most of these new technologies — and they need to be doing it now.
Katherine Strieder is global chief product officer at MiQ.