The New York Times building. Photo: Eduardo MunozAlvarez/VIEWpress/Corbis via Getty Images
The New York Times will no longer use tracking pixels from Facebook and Twitter to track its users' browser history, executives tell Axios.
What's new: The company has created a marketing tool that will allow it to target potential subscribers on platforms like Facebook and Twitter without having to leverage its users' general browsing history.
- The Times will still use trackers on a limited number of marketing pages, but it's hoping to eliminate nearly all marketing trackers in the future. It's working to make this tool work on other platforms, too.
Why it matters: "We're moving away from tracking analytics on people and towards tracking analytics on stories," says Chris Wiggins, chief data scientist at the New York Times. Wiggins says the transition will make the Times a more privacy-centric company.
"Most websites are giving up all of their users' browsing history to Facebook. The Times no longer does that."— Chris Wiggins
How it works: The new tool, called TAFI (Twitter and Facebook Interface), uses machine learning to identify which promoted articles on social media are most likely to bring in new subscribers to the New York Times when targeted to the right people.
- The tool uses machine learning to identify people's interests on tech platforms, like what articles they like on Facebook or accounts they follow on Twitter, and then targets certain articles to those people.
- Eventually, the technology will optimize social posts across all platforms, but for now it can only optimize posts within the ecosystem of individual platforms, says Colin Russel, the lead data scientist behind the tool.
Between the lines: The tool will also help the Times save money by cutting off its reliance on paying social platforms and other third-party ad tech companies for their ad-tracking technology.
- Wiggins says that the company has saved 6% of its total marketing spend by cutting out those third parties. "It's cut our CPOs [cost per order] in half when tested against manual campaigns. ... Our money going a lot further in garnering more subscriptions."
What's next: The team says it's currently building out the tool's functionality to work for other platforms, like Google search, Reddit and Snapchat.