Snowflake and governance platform OneTrust this week announced a new collaboration that brings the latter’s consent signals directly into the former’s data clean rooms, a cloud-based solution for sharing data among organizations.
The partnership looks to solve the disconnect between the consent given when data is collected and the fact that the original consent isn't always easily accessible by marketers when they use data downstream from its original collection point. The integration is a step forward for privacy-first data collaborations that are at the heart of how marketers are driving outcomes through personalization. Artificial intelligence, as it increasingly becomes integral throughout the advertising ecosystem, only complicates the issue.
“If you don't have the right consent signal going into that, when you get out, there’s a whole bunch of risk, and probably garbage data, too,” said Dennis Buchheim, global head of marketing technology for media and entertainment at Snowflake, about data collaborations.
The integration of OneTrust consent signals directly in Snowflake Data Clean Rooms involves several elements. The deal makes consent actionable across workflows; honors user choices across analytics, activation and data sharing; and scales collaboration across marketing, data and governance teams. OneTrust is used by over half of the Fortune 500 and bills itself as the largest governance provider in the world around areas including privacy, consent and third-party risk.
Adoption of data clean rooms has increased as brand, publishers and others have rushed to hoard first-party data for a variety of reasons. The addition of consent throughout the process of sharing, analyzing and activating data is the next step in forging privacy-first solutions.
“Over the last 24 months, the big driver for us related to first-party data has been people worried about signal loss, building up assets for personalization and realizing that, given that signal loss, they have to build up their first-party data stores, because third-party data stores just are not reliable anymore,” said Ojas Rege, senior vice president for privacy and data governance at OneTrust.
“Then, I think AI is this additional layer on top of it, which supercharges analytics that you can do, but it scales the good and it scales the bad. If you get something wrong, there's a lot more impact that that can have on your reputation, on your brand and so on,” Rege added.
AI as accelerant
An example of the utility of having consent signals in data is around how brands optimize digital spend. In an example, a brand deciding on which of two publishers it will spend its ad budget can use a clean room to determine which publisher has a better match to a customer base. But it is only after applying consent signals that the true number of customers that each publisher can reach is made clear. These types of scenarios around data and consent have only accelerated in the age of AI.
“AI amplifies the speed requirements. The moment that you bring new tech like AI, it amplifies any existing gaps you've got in your consent or data governance program, so to feel confident that you're activating data effectively, you want to make sure you've got the right data infrastructure in place,” Rege said.
Plus, getting data privacy wrong around AI is even worse than in previous cases, where marketers could simply delete improperly used data sets and make restitution. The remedy for having unconsented data in AI models is rolling back.
“The recourse, if you get this wrong, is hugely business impacting,” Rege said. “If I've built a whole AI model on data I shouldn't have built it on… What do I do when I have to turn it off?”
Snowflake and OneTrust are in the early process of rolling out the solution, but some early-stage clients are working on building use cases for what could be a game-changing feature in data collaboration.
“We're starting to see new use cases that, frankly, companies would not have been comfortable executing in the past because they couldn't trust that they were operating within the law,” Buchheim said. “The sky's the limit in terms of how creative they can be.”
The solution could be of use to companies throughout the ad world, including marketers, publishers, ad tech and martech firms. Companies in regulated industries and those that must be more sensitive with consumer consent could find the solution useful. But while privacy has been a topic of interest in marketing for years, the rise of AI highlights its importance.
“This notion of privacy-first or privacy by design has existed in the privacy world for a long time,” Rege said. “The time to build it in is when these data systems are being built. It's very difficult to retrofit this stuff afterwards.”
