Ethan Smith

Ethan Smith, founder and CEO of Graphite, argues that answer engine optimization is a long-tail game. Most AI prompts, in the patterns he highlights, run ten words or longer. That changes how teams brief content: vague category pages matter less than specific answers to specific buying questions.

Graphite's research also reported that AI-written content makes up only about 14 percent of what gets cited in Google and ChatGPT in their measurements. The finding challenges the assumption that flooding the web with generated pages is a reliable path to AI visibility.

Together those claims push builders toward depth and specificity. If prompts are long and citations still favor non-slop sources, then the work looks more like careful topical coverage than a volume race.

Smith's teaching roles reinforce the same long-tail thesis. Students and Reforge participants leave with a bias toward covering specific questions thoroughly, which is the opposite of the thin AI-content mills that still pitch volume as strategy.

Smith teaches AEO at Reforge and serves as an adjunct professor at IE Business School. Those roles put his playbooks in front of product and growth operators who are hiring agencies or building in-house programs.

His public interviews tend to stay operational: which prompt shapes to cover, how to measure citation, and why quality still filters what models reuse. That tone makes him useful to readers who need a program design rather than a philosophy lecture.

On the TGA 20 he sits with the builders because Graphite's research and teaching turn AEO into something a team can staff and schedule.

The 14 percent AI-written citation share finding, in Graphite's measurements, also gave quality-focused teams cover. They could point to data when declining a pure generation volume strategy.

Start with his AEO playbook interview on Lenny's Podcast. It is one of the clearer oral summaries of how Graphite thinks about long-tail demand in AI answers, and it pairs well with Graphite's published citation research for readers who want numbers next.

Long-tail AEO also aligns with how buyers actually talk to chat interfaces. People type messy, specific questions. Smith's insistence on covering those questions is why Graphite's teaching lands with product-led teams.