Rethinking Influence: Communicators and the New Rules of Generative Discovery

During a recent collaborative meeting with members of the Page community, Aaron Kwittken, founder and CEO of PRophet, led a provocative discussion on how AI, particularly large language models (LLMs), is fundamentally reshaping the way our messages are found, interpreted and trusted.

Aaron offered a clear message to CCOs: we’re not just writing for people anymore. We’re writing for algorithms that decide what people see. That shift demands a new mindset, one that brings the concept of Generative Engine Optimization (GEO) to the forefront.

Why GEO Matters

Much like search engine optimization (SEO), GEO aims to ensure brand content surfaces prominently in LLM-driven environments. But while SEO focuses on keyword matching and link-building, GEO requires contextually rich, question-driven, multi-format content that resonates with how LLMs are trained to respond. Earned media plays an outsized role here: LLMs tend to prioritize credible third-party sources, making trust and authority more crucial than ever.

The Tools Are Multiplying. So Are the Challenges.

Aaron acknowledged that while AI tools offer exciting new ways to engage stakeholders, the landscape is fragmented. “We need to treat that as investment, not cost,” Aaron said. His team uses multiple LLMs (Claude, GPT-4, Gemini, etc.), each with unique strengths, and stressed the need for communications teams to build diverse, fit-for-purpose tech stacks. This aligns with our CommTech CCO Guide, which can be accessed here.

A key takeaway: experimentation is essential. Understanding how LLMs rank sources, ingest content, and generate responses gives communicators a head start in future-proofing their strategy.

Trust Is the New Truth

Several participants raised pressing concerns around misinformation, hallucinations, and the erosion of truth online. One attendee emphasized that LLMs are only as good as the data they’re trained on, and that data quality is lacking. “We can’t build truth algorithms,” he said. “But we can build trust algorithms.”

In response, Aaron argued for a more active communications posture: “Absent algorithmic transparency, it’s on us to uphold the truth. We need to get smarter and louder.”

Other Highlights from the Collaboration Hour

• Reputation management is changing. Once something enters an LLM, it’s hard to dislodge. Traditional SEO tactics may not apply—leading some members to test “blind landing pages”; hidden FAQs that can be scraped by the AI algorithms to inform their responses.
  
• Global dynamics differ. In some markets like Indonesia, leading media outlets block LLMs from scraping their sites—limiting visibility and impact.
  
• Educating the next generation is urgent. Multiple members underscored the need to embed AI tools and literacy in higher education and onboarding.

Ultimately, Kwittken left attendees with a call to action: get closer to your CTO, rethink your publishing practices, and start treating your communications content as machine-readable fuel for trust and discoverability.

“Automation should never be the end goal,” he said. “Performance should be. Let AI make you sharper, not smaller.”