When it comes to developing a data strategy, publishers need to balance their offensive and defensive strategies, according to Daniel Hallac, chief product officer at Hearst Newspapers.

“The way we think about it is that we have two sides to it,” he said. “One is a defensive move [because] we all know third-party cookies are dying. We need to protect against it. And we think that short-term, that’s where a lot of our focus is. But long-term … we’re going to be operating in a new environment.”

The longer-term focus is an offensive strategy that leverages data to grow revenue streams. That strategy relies on optimising the experience for each user, and Hallac broke the strategy down into five essential parts:

1. Know the users. For the past several years, Hearst has been building profiles on its users, which has allowed the company to categorise them into five groups:

• Newsletter subscribers.
• Anonymous users.
• Repeat visitors (“known anonymous”).
• Registered users.
• Paying subscribers.

“We group these users and try to identify them across our properties,” Hallac said.

2. Put the users in one place. Once users are identified, the team works on taking the identities that live in multiple systems and gathering them in one place. That includes de-duplicating users and merging different identifications across multiple systems: “We need to ensure cross-domain tracking, which Apple and others just keep making harder and harder,” Hallac said. Finally, Hearst establishes security and data governance policies around users.  

3. Enrich the users’ profiles. In parallel with gathering users in one place, Hearst is working to enrich their data profiles. “There’s a lot we know about our users, such as their reading habits, devices, time of day that they engage with us. But really what we’re missing is what are their interests, who they are demographically, what are their lifestyles? What are their purchase intentions?”

Hearst is developing look-alike capabilities and models to help find the answers to those questions.  

 4. Find more users like existing users. By using data and applying look-alike capabilities, Hearst can find additional people who might be interested in the same things. “That’s a big part of helping our advertising business,” Hallac said.

5. Make it actionable for advertising. Finally, Hearst makes sure the information it has on customers is useful and can be applied in multiple ways. “We may know that we have 10,000 users who are interested in, let’s say buying a car. But by developing lookalike capabilities, we obviously can find more people who are like them and infer that they are also could be potential car buyers. And that’s a big part of helping our advertising business.”

The key thing to remember, Hallac said, is that being actionable isn’t just about advertising and targeting: “It’s for our marketing team to drive subscriptions or a newsletter sign-up. And it even goes into our content recommendation engines so we can ensure that we try to get the most sticky behaviour from our users.”

Once the data strategy is in place, Hallac said it is activated through what they call a dynamic template. That template allows them to better optimise the experience for the user to maximum business value.

“The goal is to get the right content marketing and advertising message to the right person at the right time.”

This case study originally appeared in the INMA report, The Guide to Smart Data Strategy in Media.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .