Explain It Like I'm Five: Data clean rooms

What is a data clean room?

A secure platform where companies can share and compare their customer data.

Why would you need a special platform for that?

Most of what’s shared in a clean room is personally identifiable data — email addresses, purchase data, IP addresses — that a company has but isn’t allowed to share. A clean room ensures that sensitive data brought in doesn’t come out, which is generally considered to be safe under most privacy regulations.

So what info does come out?

Imagine you make hats. One type of clean room lets you share your e-commerce data with a retail chain. You wouldn’t see the individual customers who bought your hats from the retailer, but their data could be combined with other customers to make an aggregated summary of shopping habits, demographics, and other product pages customers visited. So, you might find out that men under 40 buy a lot of your hats in the winter. Moreover, you learn that they buy hats when they buy gloves, so you decide to also make men’s gloves.

And what’s the other kind of clean room?

The most popular clean rooms let brands share data with ad publishers, be they digital giants like Google and Amazon, media outlets like La Presse, or retail media operators like Loblaw and Kroeger.

What would I use those for?

One situation would be comparing a brand’s data with a publisher’s to make sure results were as accurate and cost-effective as possible, instead of relying on estimates. But it’s increasingly being used to target ads: If you brought your hat data to Amazon, you could not only optimize your targeting based on what did best in the past, but also show ads to audience groups that have been looking at hat product pages.