In my experience, little in the online world is as misunderstood as data clean rooms. Ask a group of marketers how they will allow for media to be bought and measured and you might hear some mumbles about privacy but with little practical explanation. The truth is that in 2021, clean rooms are still a relatively fringe, underground activity. They are behind the scenes on some powerful tools but have yet to truly break into the mainstream.
For me, clean rooms are where real-time bidding was at the early stages of the programmatic era (circa 2008-2010). A little understood technical concept that few were really using or could explain, poised to take over marketing in just a few short years.
What is a Clean Room anyway?
So let’s get down to brass tacks. A data clean room is a piece of software that allows a user to manipulate user-level personal data without ever getting at that said raw data. Typically it’s a list of customers’ contact details along with details of their relationship with a company and it can also include IDs that are connected to the user. If you want to understand how many of an advertiser’s customers have bought a certain product, live in a certain area or some other form of segmentation, then a clean room can allow for that analysis to be done without the risk of exposing the user’s data.
So far so boring, it just allows for the same functionality a CRM database or even a DMP has had for years. However, the next trick is that most clean room providers allow for data from one clean room to be compared with another, again without exposing the customer data. This means an advertiser can theoretically compare their data with a data provider or publisher to see if they share the same users and in which categories. Very useful in a world where brands want to enrich their data with selected partners.
What does it allow marketers to do?
So being able to compare and collaborate with different companies’ data is great, however, this in itself is not a function to help marketers. How does this capability drive success for the most advanced marketers today and will dominate marketing in the near future? It important at this stage that we note that there is a clear difference between clean room offerings from the giants of Google and Facebook and the rest of the market.
Customer Lifecycle Data Activation
The customer lifecycle is divided into different sections. Because the clean rooms deal with customer data we will start at that end, the bottom of the very similar marketing funnel if you like.
Because clean rooms are generally connected to customer data and CRM often the first use case is to use them to drive customer marketing. The process of segmenting your customer base for messaging across an ever-growing range of publishers to turn first-time customers into repeat customers, repeat customers into regular customers. Clean rooms can connect CRM data with email systems, display demand-side platforms and large publishers including Connected TV giants, national press players and ultimately large walled garden buyers. This overlaps heavily with the Customer Data platform space which also syndicates customer data however in many cases clean rooms can increase that ability and onboard more publishers because of the privacy-preserving way they connect data. Expect this use case to develop ever more with publishers as it becomes both a more convenient and more secure way to connect data.
Purchase intent is vital for the customer lifecycle and can be found through customer marketing. But what of new customers vital for company growth. One of the problems of relying on first-party data is that it doesn’t allow advertisers to grow their customer base, you need to collaborate with others. Data Clean rooms though can allow you to collaborate with other publishers and data providers because their first-party data will include advertiser next customers.
Let’s say this again because it’s an important point, for an advertiser – a retailer perhaps – their new customers will already be engaged with many of the publishers for which they can match customer activity. So by allowing their customer audience to be turned into lookalikes by the publisher, or even by another data provider connected to the clean room, advertisers can use their first-party data to generate new prospects. It is this very data that by collaborating between publisher, data owner and advertiser datasets you can find signals that infer purchase intent, vital for successful prospecting. This is a vitally important part of modern marketing that Facebook has used to great success for many years and clean rooms open this up in a compliant fashion to all participating publishers that have their own first-party (ideally logged in) data. This is how lookalikes and new customer marketing will best be done in online video and CTV in the new era.
Whereas prospect marketing is all about converting new customers brand marketing is all about driving awareness and consideration for long term sales. Being able to target your audience can again be done with clean rooms collaborating with data providers and publishers. Crucially though, clean rooms can also assist with understanding and control of reach and frequency. TV has long been the natural home of brand marketing budgets and this is now joined with Connected TV and online video.
Many big TV advertisers are looking to techniques involving clean rooms to calculate the incremental reach of addressable video in online video and CTV when compared with traditional broadcast TV. There is a lot to discuss here yet the major takeaway is that clean rooms are already leading the charge for the most advanced brand buys and will become an increasing part of the framework moving forwards. FMCGs and other brand focussed advertisers need to pay as much attention to data clean rooms as the customer rich players in ecommerce, travel and the like.
Measurement and bringing it all together
Data clean rooms are also at play when the most advanced advertisers are looking at measurement. Being able to handle data that contains identifiers in a compliant fashion is an obvious use case and one being heavily leant on by Google’s ads data hub and Facebook’s own data clean room. Furthermore, all clean rooms connected to a company’s customer records allow for analysis of new customers to be done and by collaborating with leading publishers new models of attribution can be carried out. These techniques are only being used by a small but growing number of advertisers today. The major challenge is that Google’s clean room is best at measuring Google-connected media, ditto for Facebook with other publishers living in other clean rooms or other similar environments. There is the need to include in any measurement framework any media beyond the purview of such clean rooms.
This, in a nutshell, is the clean room challenge, they are not yet turnkey marketing solutions and require patience and expertise to turn into a proposition with true marketer value.
Yet what value! Advertisers ahead of the curve are already having huge success in this space and that success will only grow with early adopters getting ever more value as options grow and techniques mature. More and more advertisers are getting on board and including clean rooms as part of their marketing framework.