Is SEO Dead or Did a Vicious Cycle Just Make It Worse
Every few months someone announces a funeral. SEO is dead, killed by AI and drained by zero-click results, or so the headline goes. I have been hearing some version of this since I started building content strategies, and the headline always outlives the obituary. So I want to ask a more useful question than whether the patient has a pulse. The content that fills search results today is worse than it was five years ago, blander and more repetitive, and the people writing it are working harder than ever. That is the part worth explaining, because the cause is not AI and it is not Google. The industry did this to itself.
That is the part worth explaining, because the cause is not AI and it is not Google.
Is SEO dead or did it just get worse
No, SEO is not dead. SEO got worse because the industry trapped itself in a feedback loop where writers reverse-engineer the top-ranking pages, publish near-identical content, and Google's results fill with interchangeable answers. AI search, zero-click results, and AI Overviews accelerated the decline, but the sameness started with SEO long before the first chatbot shipped. The fix is information gain, not a eulogy.
That answer matters because most of the "is SEO dead" debate argues about the wrong thing. People treat declining traffic as proof that the channel died, when what actually happened is that the channel got crowded with content nobody needed. Search still sends enormous volume. Neil Patel notes Google 89% of US web traffic in 2025. The traffic did not disappear. The reasons to click on your page did.
The vicious cycle that made SEO content worse
Here is the loop, and I watch it run on real content teams almost every week. Someone wants to rank for a term. They open the top ten results, read what those pages cover, and build a brief that tells the writer to match it. The writer produces a page that says what the ranking pages already said, plus a little more so the word count clears the competition. Google sees another page covering the consensus topics, decides it belongs in the set, and ranks it. The next person to target that term now opens eleven results that all say the same thing, and the brief gets written off that. Repeat for three years and you get what we have now, which is ten pages that are functionally one page wearing ten outfits.
The mechanism is not a secret. SEO tools are built to surface what already ranks, and that is genuinely useful for understanding a topic. The problem is what teams do with it. They treat the SERP as the assignment instead of the floor. When the assignment is "cover what the winners cover," originality becomes a risk and sameness becomes the safe choice. I have sat in planning meetings where someone flagged an original angle and the room talked them out of it because "none of the ranking pages do that." That is the cycle defending itself.
Google did not design this on purpose, but its systems reward consensus, and consensus is exactly what a feedback loop manufactures. The more the search results converge, the harder it gets for any single page to stand out, so writers add length and keyword variations to compete on volume instead of substance. The content gets longer and emptier at the same time. I wrote about the symptom of this in the piece on why most blogs embarrass the businesses that publish them, but the disease underneath is the loop itself.
How AI poured gasoline on a fire SEO already lit
Then generative AI arrived and handed everyone a tool that produces the average of the internet on demand. If the old cycle made content converge slowly, AI made it converge overnight. A model trained on the existing web, asked to write about a topic, gives you a confident summary of what already ranks. That is the same homogenization problem, now available at the rate of a thousand posts a month. The search results did not just fill with similar content. They flooded.
The traffic numbers followed. SparkToro's analysis found that fewer than a third of searches click through, with roughly 68 percent ending in no click at all. Seer Interactive measured organic click-through rate on queries that show an AI Overview falling by about 61 percent, from 1.76 percent down to 0.61 percent. When the answer sits on the results page and the ten links below it are interchangeable, people stop clicking. Inc. went as far as arguing that Google itself drove the last nail into traditional SEO by answering queries directly. I read that differently. Google is reacting to a web full of redundant pages by summarizing them, because if ten results say the same thing, why send you to one of them.
AI did not kill SEO. It exposed how little of the content was worth visiting in the first place. I dug into the related fear that Google now punishes machine-written text in the post on whether Google actually penalizes AI content, and the short version holds here too. Google does not care who typed it. It cares whether the page adds anything.
Why everyone says SEO is dead when they mean SEO got boring
When a marketer says SEO is dead, they usually mean their old playbook stopped working. Find a low-difficulty keyword, write a competent page that matches the SERP, watch it rank, collect the traffic. That playbook is dead, and good riddance, because it was the engine of the cycle. Forbes framed the shift as traditional search collapsing and $74 billion in marketing budget. The budget did not vanish into thin air. It was propping up content that produced no information gain, and that kind of content is the first thing AI Overviews and large language models make obsolete.
What people call death is really commoditization. SEO is not less alive than it was. It is less differentiated. The brands that still earn search visibility, in classic rankings and in AI citations alike, are the ones that broke out of the loop and published something a summary cannot replace. Everyone else is competing to be the most average voice on a crowded topic, and the algorithm, the AI, and the reader all treat them accordingly.
How to break the cycle with information gain
The way out is not a new acronym. It is information gain, and Google has been measuring something close to it for years. Its patent on contextual estimation of link information gain describes scoring a document by how much new information it adds for a user who has already seen other documents on the topic. Read that sentence again, because it is the whole game. A page that repeats the consensus adds nothing for a reader who already saw the consensus. A page that contributes something new is the one worth surfacing. The loop optimizes for the first kind. Breaking the loop means building the second kind on purpose.
In practice that means leading with what only you can say. Original data from your own work. A customer pattern you have watched repeat for a decade. A stance you are willing to defend when the ranking pages all hedge. When I build a content brief now, the first question is never "what do the top results cover." It is "what does this business know that nobody else is publishing," and the SERP analysis comes after, as context rather than instruction. That single reordering is the difference between feeding the cycle and escaping it. The same logic is why genuine expertise reads differently to Google's systems, which I covered in E-E-A-T breakdown.
This is also where AI search rewards you instead of replacing you. A language model summarizing a topic reaches for the source that contributed the fact everyone else borrowed. Be the source. Publish the study, name the specific result, take the position. iPullRank's work on relevance engineering keeps landing on the same conclusion, that the content earning citations is the content that says something retrievable and original rather than something generic and complete. The full method for building content that survives both algorithm updates and AI summaries lives in our SEO content strategy, and the day-to-day version of it is what our content strategy work exists to do.
SEO is not dead. It got worse because an entire industry agreed to copy each other until the copies outnumbered the originals, and then AI automated the copying. The cycle only holds if you keep feeding it the same brief everyone else writes. Stop matching the results and start adding to them, and the question of whether SEO is dead stops being interesting, because you will be the page the summary quotes instead of the page it replaces.
By Katrina Kendall
Katrina Kendall
Content Strategist at Right Thing SEO, where she helps business owners sound like the experts they already are. Her focus is on translating real-world experience — the kind that lives in a founder's head but never makes it onto the page — into content that satisfies Google's E-E-A-T standards and actually converts. Before joining Right Thing, she spent six years in B2B content strategy, where she got tired of watching brilliant operators get outranked by generic blogs written by people who'd never done the work.