With third-party cookies on the brink of extinction due to privacy concerns and new regulations such as GDPR in Europe and CCPA in California, it is time to survey new ways of sharing data that are consistent with new global privacy rules, including adapting clean room protocols for the sharing and analyzing of sensitive data.
The infrastructure of user information differs from business to business. Facebook and Google have had their own for years, which they built using first-party data. Others haven’t had the benefit of their own large pool of users and have relied on existing market infrastructure. It’s time for the rest of the market to catch up and perhaps even collaborate in order to outcompete the big companies. This will allow more businesses to effectively leverage first-party data capabilities while maintaining privacy.
According to InsiderIntelligence, Google, Amazon, and Facebook currently own about two-thirds of the overall online advertising revenue share. Additionally, they have a huge amount of proprietary first-party data.
So, in the absolutely essential digital advertising space, how do you catch up with those who have the advantage of vast pools of first-person data? Collaborating with other companies in a privacy-preserving environment will allow businesses to compete, even with hard-to-anticipate changes to the industry.
In the modern business landscape, advertising can’t be seen as a zero-sum game, but instead should be seen as a shareable, collaborative space in which you and your partners share anonymized insights across digital platforms and applications.
Sharing data is not only integral to advertising. In a recent Snowflake blog post on data-driven marketing, Lourenço Mello, Snowflake’s Product Marketing Lead for Solutions, explained that sharing data is a necessity of data-based marketing and provides companies of all sizes an opportunity for a competitive edge.
“Companies are working to deliver more personalized content and superior customer experiences by sharing data with each other,” said Mello. “This collaboration with marketing partners demands excellent data governance and privacy controls. Can I share this data with an outside person, but be very prescriptive about what they can and cannot see? Privacy preservation is critical.”
This isn’t as difficult as it may sound. The way to share data while keeping privacy intact is by employing a privacy-preserving collaboration cloud technology. Gartner has predicted that by 2023, 80% of marketers with media budgets in excess of $1B will establish such tools.
Data clean room: definition and blueprint
“Think about it like a big Venn diagram,” said Cassandra Bruni, Senior Product Marketing Manager for Industry Solutions at Snowflake. “If you have a large retailer that is a big TV advertiser sharing with a media company, they can say, ‘We have similar customers, they have similar traits,’ and share anonymized information without actually passing the data itself back and forth between each other, significantly reducing risk.”
How will retail and media companies, for example, improve their marketing, advertising, and customer relations based on this data sharing? How will they be able to compete with companies that have a much larger set of customer information to work with?
Imagine all the data points each company has and wishes to share. None of the data that contains personally identifiable information (PII), such as an email address, can or should be shared.
“That’s not something that a company is allowed to do, just send customer PII data into the marketplace,” said Bruni.
This is why privacy-preserving data clean rooms are so invaluable. They make it possible for companies to derive anonymized customer insights without sharing their sensitive customer data.
“A data clean room is a safe place that allows multiple companies, or divisions of a single company, to bring data together for joint analysis under defined guidelines and restrictions that keep the data secure,” according to Justin Langseth, Technical Director for Snowflake’s data cataloging and discovery tool.
Bruni notes that although you usually hear about data clean rooms being used by separate companies or organizations, it can also be used to good effect by different divisions within a single company. Anyone who has worked in a corporation has probably witnessed data being siloed against other departments within the same company.
Consider companies that deal with, for example, food or beverages, with each of those brands operating separately even though they are legally part of the same organization. Using a data clean room, each division can collaborate without providing a copy of what they consider proprietary data.
“It is important to remember that what we talk about as a data clean room is in essence just a specific type of database,” said Edik Mitelman, General Manager, Privacy Cloud, for AppsFlyer. To really provide function, on top of the clean room you need a full-stack solution.
“How do you do privacy-preserving collaboration of data?” asked Mitelman. “You want to do attribution. How do you do attribution without identifiers? It’s not about clean rooms. You don’t have access to the IP address of the users, or you don’t have cookies. How do you attribute without infringing on someone’s privacy, without fingerprinting? If you want to do activation, how do you put people into segments without knowing who they are? These are the privacy-preserving technologies built on the clean room. The clean room is just part of the stack.”
So the data clean room is one leg of a three-legged stool. The stack is the second leg.
“The third leg, which is just as crucial, is not quantifiable [but] rather qualifiable,” said Mitelman. “It’s trust.”
When Mitelman talks about trust he is also implying zero trust. Zero trust has been defined by Hewlett Packard Enterprise as “architectures (that) recognize trust as a vulnerability” and throughout which “(i)dentity and device attestation and authentication are required throughout the network. Every single component in the network must independently establish its trustworthiness and be authenticated by any other component it interacts with, including existing point security measures.”
In such a system, “nobody, including the provider, has access to the full puzzle,” said Mitelman.
Providing your advertising enterprise with a future-proof backbone begins with establishing the kind of stability we spoke of as a “three-legged stool.”
From there a business can build an infrastructure that will allow it to stay nimble and facilitate compliance as privacy laws continue to evolve.
To learn about how Snowflake Data Clean Rooms help publishers and marketers improve their effectiveness, read here.