By Rob Webster, Global VP of Strategy
Having just read Ronan Shields amazing post about what’s in and what’s out in digital marketing felt a little uncomfortable. How is it that after so many articles on data clean rooms any confusion still reigns as to what they do and why they are important. Let’s try and end this once and for all!
Nearly all independent data clean rooms are made for using consumer data. They contain records of customers first party data, including in most cases things like name, address, phone numbers and the like as well as probably data about their relationship with the given company (what sort of customer they are). A clean room is necessary because marketers legally must not be able to access the records of individual users, the days of sending spreadsheets around as happened back in the noughties should be a thing of the past. Instead they need a platform where they can use the data safely.
The most common practical example is that when you are pushing customer records to facebook for example for targeting you are essentially using facebook as a simple clean room. The major use case of data clean rooms in marketing is replicating this functionality across multiple publishers and environments. In some ways a clean room need be no more complex than that. If you are confused about what a clean room is start here – aficionados may argue but a more sophisticated view can grow from this acorn.
In this way, CRM platforms, marketing automation, email platforms and more all likely have elements of data clean rooms in them. There is a huge overlap with CDPs too and surely one day these different technologies will merge. What differentiates a clean room from these systems is somewhat arbitrary but likely to be a) quite how privacy preserving and secure they are b) their infrastructure credentials, clean rooms are now often being used in conjunction with data warehousing and indeed Snowflake is a enterprise cloud data warehouse with clean room functionality and c) how they can be used for data collaboration (read activation).
The privacy preserving function of a clean room is vital. For a good clean room it will be technically impossible for any individuals personal data to be gleamed from utilising the data. That any marketing activity run does not expose that data.
Data collaboration is all about comparing customers from one data owner (advertiser/publisher/data company) with another. By comparing them you allow some use cases, such as seeing overlap, building a new segment (for targeting) or starting to understand attribution (did the people I served an ad to go on to buy). Targeting is the most common use case and is becoming very common in CTV. An advertiser will sync their data with the TV company and then their common customers can be identified (there is a privacy detail here discussed later) allowing those customers to be targeted, excluded or a lookalike built. Again all things that were first done in Facebook.
Data collaboration using a clean room will work best for all partners that also utilise that clean room. So one key consideration for marketers looking at clean rooms is which other partners they work with, what their ecosystem looks like. This network play is a huge part of the value proposition. Every clean room needs to be able to explain the value of their network or why that conversation is not relevant to them (Google works in its own huge ecosystem, some new platforms claim to be interoperable).
So from a marketers perspective. You use a clean room to allow your business to use first party data more effectively. To enrich it, use it to improve your media buying, use it to measure the impact of your media buying. It’s about customers, inclusions, exclusions, lookalikes. Its about applying consistency across environments safely. It’s about a new way to measure marketing. It’s about doing many of the things we have done for years but with more safety and respect for customers data.
Behind all this is a lot of complexity. How clean rooms connect safely is the subject of much technology and innovation. So too how they enforce safety, be ready for compliance, infosec and much more (ISO 9001 anybody). Too many clean rooms start talking about this before they have explained their core value proposition – likely because they are used to talking to technologists not marketers. This is vitally important to the legal teams and compliance teams but actually of little practical concern to the marketer, they just need to know that it is safe/compliant and that it works (those interested though of course are encouraged to read up).
Confusing matters is that Google and Facebook have their own clean room products that they use for a different purpose. Ads data hub for example mostly deals with cookie/ID level data for advanced analysis. Now a cookie being personally identifiable makes this a valid use case and indeed customer record data can also be safely added to this analysis. This work is important but should be separated from all the clean room chatter in the market. You might better call this space “Clean room analytics” vs the “Customer clean room” use case described above.
Crystal? Watch advocates ramp up the confusion in the comments on LinkedIn – join the conversation here.