SEO is dying, so everybody says. Google searches are down, CTRs are in the gutter, and LLMs rarely send traffic to publisher websites.
As a practitioner, I don’t want SEO to die and certainly don’t want the rules to change. But simultaneously, I think SEO’s death can’t come soon enough. Here’s how I justify that double-think and why we all should be rooting for SEO’s pseudo-death.
Organic visibility is very broken
For most of Search’s history, the signals Google used to rank content were gameable, and boy did we game them.
Before Panda and Penguin rolled out in 2011 and 2012, the web was flooded with spun articles, link farms, and keyword-stuffed pages that ranked even though they shouldn’t have.
So Google did what it does best: fix the algorithm. Naturally that made things more complicated for publishers.
This ensured that only quality, original content ranked, but it also made visibility harder for publishers.
Consider too metrics like Domain Authority and backlinks that correlate with a site’s authority and trustworthiness. Google relies on these proxies for trust (among dozens of others) since it’s difficult to verify whether new information is authentic, reliable, or simply fabricated.
[Google doesn’t rely directly on Domain Authority figures, but there’s a clear relationship where higher DAs = better rankings. ]
But these get gamed too. You can pay a grey hat on Fiverr to juice your Domain Authority, work with a link-building agency, and do ABC link swaps till your heart’s content. But that doesn’t make your site the best authority on the subject. It just means you can play the game better than others.
There’s an even bigger problem with the way we’ve measured SEO success, and this is, I believe, the core of the problem.
SEO rarely rewards content on its own merit
The best content doesn’t get the most visibility. Instead, SEO has been rewarding the ability to fund the distribution of knowledge.
Imagine you’re a B2B founder. You’ve spent years building expertise in your field. You have real findings, real data, and hard-to-fake knowledge. But to surface said expertise and get any meaningful traction, your options were
- Hire a content agency to package and amplify it
- Build an internal team to produce content volume around it.
- Hope a press release catches the right journalist on the right day (or pay to get it distributed)
- Earn coverage in an industry publication where access is often based on relationships
Granted, all of these are very doable. It is in fact how content and visibility work today.
But none of these routes are solely about the quality of knowledge you share. They are about the budget and network available to distribute it.
Wouldn’t it be nice if something changed?
SEO (as we know it) will die. Kinda.
I’m not naive enough to think the search landscape will start rewarding good content simply on its merit. If anything, the problem might even be worse now, as AI and programmatically generated content are on the up! Google’s stance is against low-quality content. But LLMs are citing them anyway!
We’ve also seen the death of SEO predicted, proclaimed, and pondered several times.
One vivid example is in August of 2010, when WIRED ran a cover story declaring thatthe web was dead
. The argument was that social media and mobile apps were pulling users away from browsers entirely and that discovery would happen inside platforms, not through search engines.
But what happened instead was these platforms coexisting with Search and even supporting each other in many ways. Last year, Google even started surfacing Reels in search.
We heard the death of SEO proclaimed again in 2015 with the announcement of RankBrain, Google’s first machine learning system applied to search rankings. We thoughtAI was about to make SEO practitioners obsolete
. But that didn’t happen.
But today’s reality of search cannibalization and LLM search is different.
Previous disruptions changed the rules of search. We still searched as we’d always done; only the requirements for ranking high changed. But this time, LLMs are rewriting HOW we search.
We now go directly to ChatGPT and it searches on our behalf.
To avoid hallucinations, these answers get pulled from live content on the web in real time. This process is called retrieval-augmented generation, and smarter people than me have explained it in greater detail. Read more about it on Wikipedia
.
We’ve also learned that LLMs use query fan-out, where they run multiple searches for any given questions. This means that search volume is theoretically higher than ever. But the visits to your website are increasingly not part of the journey.
So… what comes next?
We know that search volume is up, overall traffic is down, and top-of-funnel and many middle-of-funnel traffic sources have evaporated.
When LLMs first entered the scene, everybody was scrambling to make sense of the situation. Now, we have a slightly clearer picture of what works. The general direction is: optimize less for Google and more for LLM queries and their long-tail, high-intent nature.
- Intent-based content
- Unique and proprietary insight
- Less focus on vanity metrics and more focus on numbers that move the needle.
The bottom line is, SEO is changing again. This time the changes are more dramatic, and there’s even more that practitioners now have to consider when optimizing any website.
Unique content can now be surfaced, even for sites with low DA, backlinks, and even poor organic visibility. Before LLMs, those pages would likely be buried beneath high-authority content.
So maybe let’s be glad some parts of SEO are dying and look to the future. The thought that high-intent content with genuine value will finally start to see the light of day fills me with hope.
