Hero Profile
Joel Hughes — Founder, Paperback Labs
Media superpower: Turning messy, legacy content into a living “insights engine” that drives privacy-safe engagement and new revenue.
Origin Story
Joel entered media right as the web began reshaping everything. A computer science grad who pictured “ink and paper,” he instead found himself in IT at a publisher as the “digital department” sat like red-armor stormtroopers at the edge of the business. That vantage point forced him to learn every function—editorial, production, accounting—and to evolve fast. He bounced across roles and companies, then in 2019 a layoff nudged him toward entrepreneurship. Consulting proved noisy and exhausting—until a prospect call about CDP value unlocked the big idea: use AI to read content and infer audience intent. Joel built the MVP himself, guided by a chat window tutor. “I have to experience things completely…and then pivot quickly,” he says. The pilot landed, the product stuck, and insights were born.
Lessons & Strategies
1) Classify content, not people.
Publishers can infer intent from the page itself. “It’s way better for privacy…kind of cookie-proof,” Joel notes. Label content for helpfulness and purchase proximity; then use those signals to power on-site recommendations, segmentation, and sales narratives—without chasing users around the web.
2) Ship the thing. Then sell it.
After that prospect call, Joel asked himself, “What if I just build this as a minimum viable product…‘I made this thing. Do you want to buy this thing?’” The move from slides to software converted curiosity into contracts. Lesson: prototype against one concrete use case; let results do the persuading.
3) Pair two databases: audience + content.
When your content graph (vectors, taxonomy, “helpfulness” index) meets your first-party graph (CDP), you get smarter targeting and products: related-content widgets that actually help; chat interfaces grounded in your archive; context windows for ad and commerce plays. Paperback Labs integrates via APIs, plugs into CMS (e.g., Param1) and CDP (e.g., Omeda), and can act as either the intelligence layer or end-to-end via widgets.
4) Build resilience as a product skill.
Joel admits he’s “a late bloomer on…resilience,” but sticking through the lows mattered as much as the code. Leaders should resource the change—train staff, align incentives, and set a cadence that survives rough patches.
What he’d tell publishers:
Start with the jobs your content can do for users today. Measure “helpfulness,” not just topic. Use LLMs to extract metadata and intent at scale, then turn those signals into products that recover search leakage and create incremental revenue.
Collaborate
Collaborate with Joel!
Website: https://www.paperbak.com/
Linkedin: https://www.linkedin.com/in/joeldhughes/
Join the Movement
Subscribe for more media/data stories and practical models: https://subscribe.peopleofmedia.com
Be a guest! Take the publisher/provider survey: https://survey.peopleofmedia.com
Who should call Joel tomorrow? Heads of digital revenue seeking new products and monetization from existing content—and audience leaders aiming to boost engagement, recover page views lost to AI search, and grow first-party data with privacy-safe signals.
