Kara Chiles, SVP of product management at Gannett and the USA TODAY Network, began her career as a newspaper journalist and has since worked in nearly every medium, except broadcast television. So naturally, what drives her is not only what we communicate, but how we do it.
Chiles has always been drawn to the flow of information and how to build digital experiences that deliver it effectively. That curiosity led her to Gannett, where she began as a digital editor for USA TODAY’s Life section. But it wasn’t long before she found herself mingling with the company’s early digital product teams.
“You could say I found my tribe,” Chiles says. “We weren’t even really calling it ‘product’ yet. It was about figuring out how to take our great journalism and turn it into this amazing forum and format, and make it reachable to more people.”
Now as SVP of product management, Chiles leads solutions across more than 200 markets, and each day brings unique challenges.
“It’s solving puzzles every day,” she explains. “You never get complete information when you’re building technology. You have to try things and see how they go, but ideally, you have some proof points along the way.”
Recently, Chiles has been focused on the launch of Gannett’s new DeeperDive tool that brings an AI chatbot directly onto the USA TODAY site. Gannett created the tool in partnership with Taboola, which provided the AI underlying tech. The product launch exemplifies what drew Chiles to the media industry: sharing information and reaching people in new ways.
Chiles took AdMonsters on a deeper dive into Gannett’s tech stack.
Andrew Byrd: Can you walk me through the key pillars of your product strategy?
Kara Chiles: Under Kristin Roberts’ leadership, we’ve shifted to a flywheel approach to product strategy. While we still have core pillars that guide us over a two- to four-year horizon, our focus now centers on a continuous cycle: Create great user-centric experiences, scale them across more than 200 markets, and then drive engagement and retention.
We think in stages—create, reach, engage, retain—with equal attention on both audiences and advertisers. The “flywheel” comes from executing all of that consistently and at a high level operationally.
AB: You recently announced a deeper partnership with Taboola, including the integration of its generative AI engine on the USA TODAY site. What value does that integration bring to your readers and advertisers?
KC: We’ve been exploring AI for years in low-lift, strategic ways, so the gen AI integration with Taboola was a natural next step. It built on a strong partnership—combining our trusted content and audience reach with Taboola’s speed in tech innovation.
The goal was to create a more dynamic, search-like experience rooted in credibility. We’re now in close collaboration, refining the tool based on user behavior. Early results are promising, showing value for readers and new opportunities to engage advertisers.
AB: What are some of the early results you’ve seen so far in terms of user engagement or revenue?
KC: It’s still early days, but we’re starting to see promising signs. We’ve placed DeeperDive on both our homepage and article pages, especially knowing how much of our audience comes in through mobile article traffic. While we’re still evaluating how placement affects engagement, the initial trend is positive—engagement is moving in the right direction.
Since we’re in beta, the focus right now is on understanding what’s working—what types of prompts are gaining traction—and refining from there. It’s very much a “train the trainer” phase, where we’re learning how to guide the tool to deliver the most value.
AB: With any new user-facing AI experience, you’re also gathering insights about audience behavior.. How are you using DeeperDive to better understand those signals, especially when it comes to potential revenue streams?
KC: It’s still early to draw conclusions around revenue, but we know that engagement and retention are critical signals. The real promise of DeeperDive is how it shifts the search experience—it’s not quite a conversation, but there’s a dialog-like flow.
Users can enter a query or click a prompt and not only get a result, but also suggestions for where to go next. That structure gives us new signals on how people explore content—and how they might want to interact with it going forward.
It’s also a learning moment for our users. This isn’t just a redesigned article page—it’s a new behavior, and we’re watching closely to see how audiences adapt. Internally, it’s already sparking ideas. Every insight we get opens up a “What if we try this next?” discussion. We see it as an early signal of how audience behavior might evolve—similar to what we saw with the rise of social or vertical video.
AB: Since you brought up search—how do you think AI, especially developments like Google’s AI Overviews, will impact how publishers operate over the next several years?
KC: The only certainty is that how we work will change—the bigger question is how we choose to evolve. While AI progress can feel sudden, it often follows unpredictable cycles of rapid and slow change. We’ve been exploring this space for a while, so we’re prepared for those leaps.
At the same time, Gannett has been proactive in protecting our content and asserting the value of our journalism. We believe in building partnerships—not letting others scrape and use our work without permission or compensation. Our journalists are doing important, often difficult work, and that needs to be front and center as we adapt.
As a product team, we’re excited by the challenge of shaping how audiences might want to consume our content in the future—even in ways we haven’t imagined yet. The key is staying alert to the signals of real transformation and building with that in mind.
AB: Outside of the Taboola partnership, what are some other ways you’re using AI? Are you approaching it more for automation, optimization, or something else?
KC: We think about AI in three main buckets: creating efficiencies, optimizing experiences, and personalizing for users. One example live on USA TODAY and some local sites is our Key Points feature—those short bullet summaries at the top of articles. That used to be a manual step for our journalists. Now, it’s auto-generated within our publishing tool, with editors reviewing the results before publishing. It’s a clear case of AI helping streamline workflow while keeping humans in the loop.
We’ve also automated some data journalism feeds—allowing certain content to publish without needing manual editorial input. These use cases let us focus more on storytelling, while AI handles repetitive tasks in the background.