How Stuff Digital used AI-powered audience intelligence to grow targeted ad revenue 85%

Results at a glance

The challenge: Turning reader data into something advertisers could buy

The business's scale has always been an advantage, but its reach was becoming less valuable to advertisers. About 17% of daily visitors were logged in. For the other 83%, after earlier shuttering their DMP, there was limited visibility into who those readers were or what they cared about.

The data was there. Behavioral signals, subscriber data, and content engagement all existed, but turning that information into useful audiences was difficult.

The existing setup created a few key limitations:

• Audience creation was rigid. The business had relied on Adobe Audience Manager, but the setup became difficult to work with and hard to adapt as needs changed.
• The audiences lacked depth. A sports audience might be based on little more than a few recent article views, and it was difficult to explain what separated one segment from another.
• New requests moved too slowly. Anything outside the preset audiences in Google Ad Manager often went back to the data team and required days of manual work. With briefs frequently due the next day, that sometimes meant missing the chance to respond with a tailored audience.

As the advertising market became more competitive, the need was clear: make more of its existing data available for audience building, without routing every new request through the data team.

The breakthrough: Making more of Stuff's data usable

Stuff Digital already had the foundation it needed in BigQuery, where its behavioral and subscriber data lived. What it lacked was an easy way to turn that data into audiences.

Hightouch's Agentic Marketing Platform works directly with the data in Stuff's warehouse, so marketers can build audiences using signals already available in BigQuery and activate them in Google Ad Manager without first moving that data into a separate audience platform.

Instead of relying on a few simple rules, the team could combine signals like content topics, specific entities, recency, frequency, device type, and engagement to build a much richer picture of reader behavior. A broad sports audience, for example, could be refined based on the sports someone follows, how frequently they engage, how recently they visited, or the device they use.

The team calls these richer segments smart audiences.

The difference was immediately noticeable to the people building them. Caroline remembers one member of her team "almost exploding with excitement" when he saw how easily he could work with the platform and its AI agent.

Adding intelligence to the audience

With Hightouch's Agentic CDP, the team could go beyond building more precise audiences and use AI to understand, refine, and uncover new opportunities within them.

While reviewing its sports audience, for example, the team asked the agent whether there were relevant signals they might be missing. It suggested New Zealand golfer Ryan Fox, who is well-known among local readers and had just that week gained global attention by winning the British Open.

The agent can also analyze an existing audience and surface insights about the members. Those insights can go directly into advertiser pitches, giving sellers a clearer picture of the behaviors and interests that define the audiences.

Previously, producing that kind of analysis would have required significant analyst time. Caroline says that type of analysis simply wasn't happening before. Now, those richer audience insights are part of the sales conversation.

The results: A more valuable product for advertisers

Stuff Digital paired the technology change with a new commercial approach.

The sales team was trained to move conversations away from inexpensive inventory and clicks. Instead, they pitched reaching the right audience and driving better business outcomes.

With a more precise, differentiated audience product to sell, the team launched the new targeted audiences at roughly 2x the rate of run-of-site inventory and held that pricing without discounting.

Advertisers bought them.

In the first three months, targeted advertising revenue was 85% higher than during the same period the previous year. Early performance data was promising, with smart audiences generating a 41% higher click-through rate than general ads on the site.

The business case moved faster than expected too, with the business reaching its six-month return target in just three months.

The impact also shows up in how quickly the team can respond to an advertiser. In one recent fast-turnaround pitch, John shared a brief with the team. About 10 minutes later, they had identified three relevant audiences representing roughly 850,000 people, along with additional audience intelligence they could take back into the pitch.

A new era

For John, Hightouch's Agentic Marketing Platform gives the team a stronger position in the advertising market. The company is now taking its targeting capabilities on the road and having more sophisticated conversations with advertisers about audiences and outcomes.

For Caroline, the difference is what her team can say yes to.

A year earlier, new ideas for using Stuff's data often ran into the same underlying constraint: the data existed, but getting it into a usable form was another matter.

Today, that foundation is in place.

What's next

So far, the focus has primarily been on the advertising business. Next, its audience growth team plans to use the same foundation to build audiences designed to bring readers back and increase ongoing engagement.

As Stuff Digital adds new data and signals, its audiences and the insights built on them will keep getting sharper.