Danny Sullivan on SEO – “Optimising for Humans” Still Wins
Google’s Danny Sullivan recently addressed the ongoing debate around SEO and AI-powered search, making it clear that Google’s ranking systems are tuned toward a single goal, regardless of whether results come from classic search or AI experiences. That goal hasn’t changed: satisfy people.
The Search Engine Journal is reporting that it’s a message the search industry has heard before, but as AI reshapes how search looks and feels, it’s one that’s likely to dominate SEO conversations even more in the years ahead.
Is AI Search a Fad?
During the discussion, Google’s John Mueller asked whether AI represents a genuinely new shift or just another passing trend:
“So everything kind of around AI, or is this really a new thing? It feels like these fads come and go. Is AI a fad?”
Sullivan’s response cut through the noise. He joked about the growing list of acronyms — GEO, AEO, AIEO — and expressed relief at not having to chase every new label. More importantly, he emphasized that despite changes in search formats, there isn’t much that site owners need to do differently.
Search experiences may look new, but the underlying expectations have not. According to Sullivan, Google has repeatedly consulted its engineers to answer the same question from publishers — What should we be doing now? — and the conclusion has remained consistent: nothing fundamentally new.
What Google’s Systems Are Actually Designed to Rank
Sullivan then shifted to what Google’s ranking systems are built to reward. His comments aligned closely with recent remarks from Robbie Stein, Google’s VP of Product for Search, who explained how human feedback helps train Google’s systems to recognize helpful content.
While Sullivan didn’t dive into specific signals, he reinforced the same core idea: Google’s systems are engineered to identify content that satisfies humans.
He acknowledged why people assume they need to change tactics when they see new AI-driven search experiences:
“People really see stuff and they think they want to be doing something different. It’s a natural reaction.”
But he stressed that Google’s “North Star” hasn’t moved.
Content for Humans, Not Algorithms
Sullivan was explicit about what not to optimize for:
“All of our ranking systems are about rewarding content that’s great for people, written for human beings, not written for search algorithms, not written for LLMs, not written for whatever acronym you want to use.”
Google’s improvements, across classic search and AI-powered results, are all aimed at better identifying content that users find satisfying and useful. Trying to optimise for specific AI systems, Sullivan warned, risks pulling creators away from that central goal.
As those systems improve, content designed to “game” them will likely fall behind, forcing publishers into constant catch-up mode.
Why Optimising for LLMs Is a Mistake
Although Sullivan didn’t cite traffic data directly, the implication is clear. AI tools like OpenAI, Perplexity, and Claude currently drive less than 1% of referral traffic combined. Optimising for large language models at the expense of search visibility is a losing trade-off.
Content that genuinely helps people remains the safest, and most scalable, strategy.
Why SEOs Have Been Skeptical
Google has claimed for decades that its algorithms prioritize user satisfaction, and for a long time, many SEOs were skeptical. That skepticism wasn’t unfounded.
However, since at least the 2018 Medic core update, Google has made measurable progress in aligning rankings with real user value. Advances in AI, neural networks, and large-scale human feedback have made it easier for Google to understand what people actually want from search results.
Robbie Stein’s recent explanations of how aggregated human feedback shapes rankings underscore how far Google’s systems have come.
Is Human-Optimised Content the New SEO?
Links are no longer the dominant ranking factor they once were. Google now excels at understanding queries, interpreting content, and matching the two based on user satisfaction signals, many of which have existed in some form since the early 2000s.
This raises an uncomfortable but necessary question for the industry: is it time to abandon old SEO playbooks?
If Google’s systems are increasingly aligned with human preferences, then “optimising for humans” may no longer be a slogan, it may be the most effective SEO strategy left.
For SEOs and creators alike, the future looks less like chasing algorithms and more like building genuinely useful experiences for real people.
