SEO ‘Is on A Dying Path or Not”: What Google Says?

As usual, Google conducted its Search Off the Record Podcast recently.  When the title was “wrapping up 2024”, the main theme revolved around the impact of AI on SEO.

All four hosts on the podcast discussed the implications of generative AI on SEO.

When asked, Garry clearly negated the notion that SEO is on its dying path.

Sarcastically, he said it had been happening since 2001, according to the Internet.

He even said, “I expected the very first article of 2025 to be about how SEO is dying again.”

In response to this statement, another host requested a counterpoint regarding the relevance of SEO in 2025.

Is SEO still relevant in 2025?

While addressing Lissi Sazman’s question about SEO’s relevance in 2025, they referred to RAG or Retrieval Augmented Generation.

You already know that LLMs play an important role in Gen AI. These trained models are responsible for generating any kind of information out of their data pool. However, sometimes, they make mistakes or misinterpret, resulting in producing wrong information or what we call AI hallucinating.

That means they are not perfect. So, RAG or Retrieval Augmented Generation is the solution to minimise this drawback of LLMs.

What is Retrieval Augmented Generation?

RAG is a technology that helps LLM retrieve information from external sources or databases, such as crawled or indexed websites, to generate information more accurately.

Now that the LLMs don’t rely merely on their own data but cross-check information from external resources, their chances of hallucinating decrease.

Martin Spilt explains this with examples of producing something out of the bag of documents.

Documents represent all those search indexes or knowledge bases from which LLMs retrieve information to produce the output result for any query.

To be more precise, in Martin’s own words:

 “Probably nowadays it’s much better, and you can just show that, like here, you upload these five documents, and then based on those five documents, you get something out of the bag.”

Connection Between RAG and SEO

Continuing with Martin’s document analogy, John discussed how SEO can be connected with RAG.

He said:

I found it useful when talking about things like AI in search results or combined with search results where SEOs, I feel initially, when they think about this topic, think, “Oh, this AI is this big magic box and nobody knows what is happening in there.

And, when you talk about kind of the retrieval augmented part, that’s basically what SEOs work on, like making content that’s crawlable and indexable for Search and that kind of flows into all of these AI overviews.

So I kind of found that angle as being something to show, especially to SEOs who are kind of afraid of AI and all of these things, that actually, these AI-powered search results are often a mix of the existing things that you’re already doing. And it’s not that it suddenly replaces crawling and indexing.”

Simply put, John Mueller is saying that AI is still evolving and isn’t completely perfect, and to make it better, LLMs have to rely on RAGs.

For RAG to work, there should be websites or enough optimised content on the Internet that RAG crawls.

Because crawling and indexing are happening, SEO is still relevant in 2025.

He is simply trying to say that you must not ignore your website structure or SEO just because AI plays a big role in search rankings, as discoverability and crawlability are still things that keep you active in the eyes of search bots.

Related article: Optimising For Google’s New Landscape and The Future of Search

Our Thoughts

When John is correct that SEO is still relevant for crawling and indexing, that doesn’t change the fact that AI somehow impacts rankings in SERP.

Whether AI algorithms decide rankings or AI Overviews, both have somewhat negatively impacted organic SERPs.

Firstly, SGE has taken a significant portion of organic results, shifting them a little downwards. Second, these snapshots give quick, concise, and precise information in one instance, so people usually don’t bother clicking on blue links unless and until they want to dig deeper.

Moreover, this scenario has also impacted SERPs. For continual information, SGE has a follow-up option, which again reduces the number of users heading to organically indexed pages to some extent.

Another factor impacting this is the rise of NLP queries. AI has made people fond of seeking information in a flow, like smoothly answering a question, which was earlier mostly limited to specific keyword searches. Natural language processing ability precisely answers any question, no matter how long it is. When these queries are possible to target by opting for long-tail keywords, AI overviews better handle them.

Now, let’s get to AI algorithms influencing search rankings. When Google has accepted that these algorithms consider many ranking factors while placing results in order, there is opacity about how they exactly work behind the curtains. Google has indeed revealed many of them, like EEAT and YMYL, etc., but still, there isn’t much transparency that SEO experts expect.

In short, while it is true that AI doesn’t affect SEO until it concerns indexing or crawling, it directly or indirectly impacts the overall organic SERP rankings, even if not completely killing SEO. So, it seems that the only way to overcome this impact is to adapt SEO strategies to new demands and trends.

Thanks for reading. If you find the information helpful, please subscribe to our blog. Here, you can get more valuable information on digital marketing, AI, SEO trends and more.