Data Clean Room:
What It Is & Why It Matters in a Cookieless World
Generally speaking a Data Clean Room (DCR) is a cloud-based infrastructure and platform in which both advertisers and publishers, as well as Third-Party data providers, can ingest their respective sets of First-Party data, without the other parties having factual access to their partner’s data. The data of each contributor is then being processed by the platform, with the purpose of finding patterns or overlaps within the whole data pool. The currently most significant use case that can be implemented with the help of a Data Clean Room would be advertising performance measurement, where the advertiser provides first-party data of contacts that have made a certain conversion (e.g. a purchase, a visit to a campaign landingpage, etc.) which is then being matched against a set of first-party data coming from a publisher, that has shown complimenting ads to the conversion. The intersection between these data pools then represents the number of contacts for which the ad appears to have worked. Further insights like the performance of individual motifs within the pool of ads for a campaign or insights on ad frequency and respective contact reaction (e.g. a Click) can be utilized for planning future campaigns and optimizing both dynamic creative content as well as contact frequencies.
It might also be worth pointing out, that Data Clean Rooms cannot be utilized for direct access to any other than the company’s own data. During the configuration stages of a DCR, a business needs to link their specific advertiser accounts to the Clean Room, and these are the only data sources that will provide the business with access to its raw data, based of course on the data source’s own connection and access restrictions.
It is also good to know that their a different types of DCRs which differentiate themselves through different setups and operating models.
First of all, there are Media Data Clean Rooms. These DCRs like Google’s Ads Data Hub (ADH), the Amazon Marketing Cloud or up until July 1st, 2021, Facebook’s Advanced Analytics will provide the marketer with more options to evaluate their campaign performances within these platforms, to build better audience segments and to provide these audiences with more meaningful content.. This will ultimately result in an overall better, more optimized ad spend within this one platform! That is also the single biggest drawback of these DCRs: They are only working with and considering data which is being used within their own platform. You cannot ingest analytics data from your Amazon shop into Google ADH for example, you cannot access any kind of data from across other ad platforms. Considering this, Media DCRs are a good choice for any marketer who already spends most of its advertising budget on the specific platform but Data Clean Room technology in general will not be the solution to finally breach into these walled gardens to freely access its data.
The other kind of DCR that we can see being established are Partner or Distributed Data Clean Rooms, with vendors being Nth Party, Kochava, Infosum, LiveRamp and Snwoflake, amongst others. These DCRs can be used by any partners like marketers, publishers, agencies or Third-Party data providers to safely share their data assets with each other, making use of multi-party encryption technologies and data science. Every partner maintains it’s data ownership and can apply strict controls on how much of their data is being shared, with whom and how it can be accessed. This operating model differs significantly from the one put in place by the Media DCR vendors, as the partners can freely define their own set of rules and regulations when it comes to sharing and analyzing data or enriching it with datasets, graphs or integrations with other sources. However, the trade-off for having this much control, granularity and flexibility over the data assets is getting significantly fewer data points from the walled gardens themselves.
And while this problem can be resolved by operating Media and Distributed DCRs in parallel, each for their respective use cases and platforms, that would still prevent the marketer from getting a thorough, holistic picture of their data assets and it’s activation potential. Scalability of the platforms can also be some kind of an obstacle to overcome during the initial phases of setup and operation, as this heavily depends on how much partners can be found and to what extend those are able and willing to participate in a concept of open and free data sharing within Distributed DCRs. But things are not just complex and complicated, there is also some great value to be had when putting a DCR to work for your organization, even at this early point in time with regards to technology maturity and establishment.