Back to Pages

Data Clean Rooms: Everything You Need to Know

  • 12 min read

What is a data clean room?

In scientific research or manufacturing, a “clean room” is a strictly controlled environment that reduces the level of contaminants that can put sensitive work at risk. A data clean room is very similar, and these digital spaces are now growing in popularity due to organisations embracing digital transformation.

So how do we define a data clean room? Simply put, It is a safe and neutral space for data collaboration and partnerships to exist without either party (or parties) having access to the other’s customer data. A data clean room also embeds privacy-enhancing technologies, such as encryption and differential privacy, so data can’t be used inappropriately while providing data scientists with the ability to leverage data to better plan, activate, and measure across the ecosystem. 

Think of a clean room as a worthwhile addition to the data transformation toolbox. One of the key benefits is that the data never leaves the data owner’s control, creating a better balance between privacy and utility. While many use cases are between two enterprises, data clean rooms can also be used across the same enterprise for the same reasons.

History of data clean rooms

When the industries of adtech and martech think of the origin of data clean rooms, they point to the introduction of Google’s Ads Data Hub (ADH), with Facebook and Amazon following shortly after. These entries into the space were products designed for secure data collaboration among the walled gardens and advertisers (and their first-party data) to enrich, target, and measure campaigns on their platforms. It also enabled these walled gardens to avoid risking the exposure of their coveted consumer data. Enhanced privacy and security regulations continue to make campaign activation, measurement, and optimisation without safe data collaboration in a data clean room extremely challenging, if not impossible. Stricter privacy and security laws around data and changes from browsers and device manufacturers have made data clean rooms more of a necessity in today’s world rather than a best practice.

How do data clean rooms work?

There are essentially four components to a data clean room.

  1. The data: First-party data comes from participants like a retailer and CPG.
  2. Data connection and enrichment process: This is where the data sets from two or more parties are matched on an individual basis and then enriched using third-party vendors. It’s also where an identity graph, which will pseudonymise to an ID that can’t be tied back to a consumer’s PII, comes into play.
  3. Data analysis and analytics: Within the data clean room, anonymised data, which can be used for measurement and attribution, is onboarded.
  4. Marketing Activities: The last and most important component is using the output from the data clean room to execute marketing activities, such as audience building, customer insights, determining reach and frequency, campaign analysis, customer journey analysis, and more.

Data Collaboration and Clean Rooms

How Data Collaboration Can Take Your Clean Room to the Next Level

GO TO VIDEO

Types of data clean rooms

As we’ve noted, data clean rooms aren’t new. The walled gardens have had them for several years and have proven their effectiveness at improving advertising performance through data collaboration between an advertiser and its platforms. This approach helps advertisers check for inconsistencies to address issues such as frequency capping and audience suppression, preventing over-serving ads to the same people or households. With the proliferation of channels, measurement and attribution become even more critical to optimising advertising spend.

Neutral third-party data clean rooms

More data clean rooms are emerging. In addition to those managed by the platforms, there are also neutral third-party data clean rooms, like LiveRamp’s Safe Haven, that, in addition to organising, analysing, and measuring data, can also enable activation across the entire advertising ecosystem to better plan, activate, and measure your marketing spend. 

Having a safe place for your data scientists to experiment is great, especially if it’s where the data already lives. However, data clean rooms are only as valuable as the value they produce. Science is all about testing and providing proof points. Data science should improve efficiency, power the discovery of new audiences, and increase accurate measurement. An enhanced, neutral environment is better suited to do so, especially when working across multiple platforms and channels.

Data clean rooms for better business outcomes

Today, data clean rooms can provide a safe and secure environment to unite disparate data sets more easily and uncover more business insights than ever. While data privacy concerns are a solid reason to use a clean room, there are other benefits that will help your businesses achieve more from your investment in the long term — that is, if you’re choosing the right type of clean room. After all, the ultimate goal is to create value for your business.

To achieve better business outcomes, your data clean room must be rooted in privacy protection and provide the tools that deliver results for marketers. An enhanced data clean room with embedded people-based identity, measurement, and insight applications that are also connected with leading activation channels can help you build deeper partnerships and network your sphere of influence.

Clean Room Primer

Everything you need to know about clean rooms

DOWNLOAD NOW

Three reasons marketers should love data clean rooms

  1. Accessing more (and better) data to improve marketing and maximise ROAS

As mentioned earlier, data deprecation is real, and marketers require campaign data that doesn’t rely on signal losses. With a data clean room, marketers can tap into attribution models leveraging transaction data combined with a publisher or partner’s ad performance, like impressions and clicks to determine the source of conversions. With granular data, a marketer can analyse the reach and frequency of its campaigns and take measures to minimise ad fatigue and improve media performance. As the media landscape becomes more fragmented, a data clean room can help overcome the obstacles each new or deprecated channel brings.

  1. Audience insights to build better audiences

User-level data is gold but has become increasingly harder to obtain due to much-needed privacy regulations and changes from Apple. A data clean room enables marketers to partner with multiple trusted sources to enhance and enrich their data to create look-alike models on a granular level. This can then power segmentation based on consumer behaviour, purchase history, habits, etc., to help marketers deliver more relevant messages via the most effective channels.

  1. First-party data partnerships

Egg and bacon… Ant and Dec… Ticketmaster and Spotify. Three examples of things that just go together and make sense. With a data clean room, a marketer can partner with anyone in a safe and permission-based environment to propel its brand in previously impossible ways. For example, if you’re getting ready to launch a new product, the right data partnership and datasets can help you understand the category better and drive better outcomes. First-party data partnerships benefit not only the marketer but also product development, customer service, and other teams across an organisation.

What to look for in an (enhanced) data clean room

So how can you find a data clean room that delivers these outcomes and also better customer experiences? Here are the “musts” for data clean rooms.

A data clean room must have a deterministic omnichannel approach to identity.
Enhanced privacy capabilities are often the first reason to gravitate toward a data clean room. Privacy-enhancing technologies (PETs) enable companies to analyse data without it having to be exposed. However, the key to success for data clean rooms is to provide modular privacy-enhancing technologies so customers can leverage the required level of privacy thresholds when collaborating with their partners. In addition, having a deterministic approach to identity and matching is also key to ensuring better collaboration. Without deterministic IDs and matching, the outcomes of collaboration can lead to inaccurate results. 

Leveraging an interoperable identity framework can help you not only measure effectiveness more accurately but also drive personalised experiences across omni channel touch points. When it comes to measurement, the right data clean room provides similar capabilities as walled gardens, such as: 

  • Matching impression data to transactions to deliver conversion and sales lift reporting, 
  • Enabling multiple partnerships for multi-touch attribution, and
  • Incorporating TV partners to plan omnichannel campaigns better. 

A data clean room must answer the right questions

Enhanced or not, a data clean room would be moot if it didn’t enable easier data collaboration within the privacy-conscious space. Once the data is accessible, clean rooms should also provide additional capabilities to make collaboration easier and simpler. These include:

  • Helping build and reconcile an organisation’s own first-party data sets to support first-party graphs. 
  • Ensuring organisations are bringing data sets into a unified taxonomy to support easy collaboration. 
  • Pre-building queries across data sets to answer key business outcome questions without a heavy lift from both parties.
  • Embedding analytics dashboards to ensure granular data can be easily leveraged to create better audiences at scale.
  • Embedding measurement protocols to automatically and continuously measure the effectiveness of collaborative campaigns.

A data clean room must enhance personalisation.

Data is the key to great customer experiences, and an enhanced data clean room helps unlock those experiences.

Let’s use frequent visits to your neighbourhood coffee shop as an example. By the end of the first or second week of daily visits, you would likely expect a more personalised experience from your regular barista, given he or she would know your specific preference for coffee. If nothing else, you would hope to be recognised as an individual vs. a random customer. You might even get a pastry on the house as a thank you for being a loyal customer. 

Another example is a retailer with transaction data that tells you what customers are buying, how much they are buying, and how often. That retailer could build audiences that are attractive for CPGs who may be first-party data poor. This is a win-win scenario for both parties since the brand would achieve better targeting, reach the right audiences, and reduce advertising waste. Meanwhile, the retailer (in this case, acting as a media publisher) could improve yields and deliver better customer experiences.

When appropriately leveraged in a privacy-safe manner, data can enhance the customer experience, ensuring that brands can recognise these preferences and deepen relationships with existing customers for increased brand loyalty and awareness. A data clean room delivers the ability to create exceptional customer experiences by helping activate these insights across marketing applications.

A data clean room must be flexible and interoperable

An enhanced data clean room should make it safe and easy to connect data wherever it lives and use whatever tools the marketer needs to achieve its goals. Data clean rooms should be able to work where you (and your partners) work. This means they should be able to plug into the same clouds and walled gardens. Things can get complicated when there are point solutions with different setups and parameters used by a brand or agency with multiple partners.

How retailers can monetise data and create revenue streams with a data clean room

Suppose you already have your data structured and taxonomised, and you want to make it available so that you can start generating revenue. In that case, you don’t need a data clean room because you can push your audience segments through different data marketplaces. The caveat is that you are competing with at least 150 data providers trying to go after the same media dollars. This off-the-shelf capability doesn’t provide the right kind of value for your own data sets. To improve the value of your data and start generating incremental high-margin revenue streams for yourself, a clean room can really help. 

For example, most retailers today have a retail media business focused on owned and operated properties. One of the main objectives is to improve yields on a retailer’s existing inventory, and the best way to do that is by leveraging a clean room and allowing CPG brands to do overlaps and identify unique audiences that they would like to activate across a retailer’s owned and operated web properties. That extra step allows a retailer to bundle its data and media together with the use of a data clean room. This increases the value of its inventory and owned and operated operations.

Further, to provide audience extension (where a retailer is taking its data sets and making them available or using them outside their own four walls), having access to clean rooms can provide very unique capabilities, which include both simple and advanced audience building capabilities along with measurement, which CPG brands covet.

Data clean rooms aren't just for retailers

While retailers and brands are great examples of how data clean rooms can create better business outcomes, just about every vertical can benefit from leveraging a data clean room with the right partners. Here are a few examples:

Utilities: Consider the growing market for new smart-home energy products and the data utilities have on consumer consumption. This gives them a competitive advantage for creating and marketing a variety of products and services. For example, a data clean room can enable collaboration between the utility and electric vehicle (EV) manufacturers to determine optimal charging point facilities.

Health care: The greatest and most current example of how data clean rooms can help in the health care sector is their use to provide critical information to the medical community, governments, and the world population during COVID-19. Data clean rooms strike a balance between public health and privacy with data collaboration without exposing individuals’ private data.

Fintech: When it comes to financial data, privacy and security are two of the most significant issues that face the industry. The level of PII and fallout from a data breach can be devastating to both individuals and institutions. Data clean rooms can provide a safe and secure environment for this industry to reduce risk when collaborating internally and externally. They can also be used across the industry to weed out fraud.

Entertainment: From streaming services to cinemas, concert and sports venues to theme parks, and beyond — this industry can capture consumer behaviour and purchases in ways other industries can’t. A data clean room can help unlock the value of their data through partnerships with music labels, production studios, etc., by providing unique audiences and signals for better planning, activation, and measurement.

Travel industry: Hotels and airlines have unique customer data at scale and have already embraced media networks and data clean rooms to increase revenue and partner with brands to enhance their offerings, differentiating themselves from their competition.

Mobile apps: Mobile publishers face a lot of competition for time and eyeballs, from app awareness to install and engagement to long-term stickiness. As with many mobile apps, the potential for the amount of data earned from active users is significant and extremely valuable. Consider that mobile publishers often have several different games with user activity across two or more of them. This increases the amount of data that can be used to draw insights and provide areas of optimisation, and, when safely shared with partners via a data clean room, it can bridge on- and offline worlds.

Why aren’t data clean rooms used all the time?

If data clean rooms offer so much for marketers, why aren’t they more widely used? One reason is that while they are a hot topic now, they haven’t been around for long. Other factors contributing to the adoption of a data clean room:

Getting data in a usable state

In order to use a data clean room, you need to get your data house in order. That’s not a quick or easy task for many organisations, especially if that data lives in silos.

Finding the right partners

Choosing a data clean room and partner(s) to work with can take time even if you have full buy-in from various members of the organisation. A data clean room isn’t just about or for marketing. 

Privacy and security concerns

Data clean rooms are all about data collaboration. Not every organisation will feel comfortable revealing data that may be proprietary, such as transactional data from online and offline channels. Educating different governance teams can take time, and as privacy-enhancing technologies become mainstream, the privacy and security concerns will reduce.

Restrictive

Most brands’ data science teams will be comfortable navigating the tools needed to power analytic tasks, from lead scoring to journey modelling to channel performance measurement. Since data clean rooms must create a protected environment to enable the discovery of needed insights without allowing the exposure or re-identification of consumer data, it creates restrictions that often prevent access to other tools in order to manage the safety of the data analytics.

The future of data clean rooms

Organisations will increasingly adopt data clean rooms in the coming years, but the technology will evolve with customer requirements. As we already see, customers are migrating their data assets within their cloud data warehouses, and customers will want the collaboration to occur within their data warehouses’ confines. Data clean rooms will become an integral part of the cloud infrastructure going forward and deliver value not only for marketing use cases but use cases to support all enterprise functions. In addition to becoming embedded in the cloud, we expect the clean rooms to provide modular privacy controls so organisations can feel comfortable with the privacy and security frameworks used in collaboration.

In conclusion

No matter which type of data clean room you choose for your data collaboration needs, it’s fundamental to ensure that it will help you deliver the best business outcomes without sacrificing the deep customer relationships you seek to grow and solidify as you develop a better understanding of your customers.