SEO is dead. At least that’s what doom-mongering industry commentators like to say.
I’ve been in SEO professionally since 2011 and have seen huge changes to Google’s algorithms and how SERPs (search engine results pages) are presented.
From Google’s early spam algorithms such as Panda, to more recent algorithm improvement programmes such as its regular core updates, alongside machine learning and NLU (natural language understanding) initiatives such as RankBrain and BERT; SERP volatility and a dynamic approach to SEO has become the norm.
It’s not just pure algorithmic changes we’ve had to deal with either. In my first month as a professional SEO, way back in October 2011, Google announced that SSL encryption of search queries would be the default for logged in users. That meant, instead of full keyword data being reported in Google Analytics, organic keywords for logged-in users would come through as (not provided), essentially losing the vast majority of keyword data we could previously see.
In 2013, this became the default for all users and all of a sudden, the way of reporting on SEO performance needed to change completely. At the time, it felt like a major mountain to climb. Many working in SEO now probably don’t remember being able to see exactly which keywords resulted in leads or sales but believe me, it was bliss from a strategic perspective.
But, we’re now in 2025 and I should be a hardened SEO practitioner, ready to tackle any new challenge Google throws at me. But in 2025, we now have AI as a daily part of our lives. The number of people I know in my personal life, some of which require my assistance to use a TV remote effectively, are using ChatGPT to get instant answers to all of the burning questions they have about the world.
ChatGPT is clearly a threat to Google whose main revenue income is through advertising. Any drop in search volume means less people to show ads to, which means fewer clicks and therefore less income. When people search for information in ChatGPT, they’re not doing what they would have done before, searching for that information on Google. Google needed to respond.
It did so initially with Bard, a ChatGPT-like AI chatbot which sat outside of the search environment. Eventually it was rebranded as Gemini and now has some premium features with Gemini Advanced.
Alongside the development of Bard, Google was also working on showing AI overviews in search results – the AI driven panel that you see at the top of the results, normally when asking questions. This was rolled out in 2024 and has been with us ever since.
From a user perspective, you can see why this feature is helpful – it gives you quick answers to your questions and is likely to be trained on the information found on the websites you would have previously visited to get your answer, therefore saving you time. But this is the inherent problem for publishers – their organic search traffic levels are dropping because users no longer need to click through to their sites to have their questions answered.
I use the term ‘publisher’ loosely here. I’m not just talking about news websites, I’m talking about brands who have essentially become publishers through an extensive investment in content marketing. Creating a programme of expert-led informational content is a great way to appear, at different stages of the marketing funnel in search, in front of prospective and existing customers.
Agriculture is a research heavy industry and farmers use Google all the time, so we’d always recommend that this sort of content programme should feature as part of the agribusiness marketing activity mix. However, we have seen the impact on click-through rates and traffic since Google’s AI overview panels have been introduced.
A recent study by Ahrefs showed that for informational queries, AI overviews reduced clicks by 34.5% which is massive.
Google has now upped the bar though and rolled out, in the US for now (but it will, sure as night follows day, roll out to the rest of the world in the coming months), AI Mode – a tab at the top of the search results which allows users to enter an AI chatbot type screen where they can interact with the AI conversationally to find the information they’re looking for.
Naturally, people in the SEO world are concerned, as are the brands engaged in search-driven content marketing. The fear is that users no longer really need websites when looking for information, they can simply converse with the AI to surface everything they need – it’s faster, more convenient, and will potentially give them more value due to them being able to ask additional questions in a conversational way.
Having seen what’s been rolled out so far, here’s a quick summary of our findings:
Clearly Google is including shopping ads in the results which makes complete sense as it isn’t going to want to lose ad revenue on those if the uptake is big.
From a publisher perspective, it feels like bad news. There will be cases where a user will want to click on one of the website listings to read more information about a particular thing but, in the main, I feel that users will spend most of their time conversing with the AI to get very specific answers.
Google has said that data from AI Mode will be included in Search Console but, at this stage, you can’t filter to see only AI Mode data or alternatively, filter it out to only look at normal search result data. I suspect that it might change over time but right now, impact analysis is likely to be difficult.
A lot of people, including ourselves, when analysing Google Search Console data to see the impact of AI overviews are finding that impressions are going up but clicks are going down. This makes sense because, when a website is used as source, it will count as an impression, but if the AI overview fully answers the question then the user has little reason to click through to the site. I suspect AI Mode will supercharge this trend.
So, what should publishers and brands that create informational content do? Is it time to throw in the towel and divert budget elsewhere? I’d recommend against any knee-jerk reactions at this stage. Here’s why:
As it stands, it’s only live in the US and we don’t know (only Google does) what the uptake has been like.
Once fully rolled out, not everyone will want to use it and those who do are likely to use it in different ways. We might just need to accept that traffic levels will be a bit lower to informational content – the important thing is that it’s still viable and your tracking can still prove valuable.
Google doesn’t stand still. AI Mode now will look different to AI mode in 12 months. We don’t know at this stage how Google will introduce more ad formats to it for example.
I’ve broken this section down into five key components of what I feel should be the priority for publishers moving forward. These should position brands that wish to continue with a search-focused content marketing strategy in the best way possible.
When I’m searching for information online, especially within a topic close to me (like a farmer when searching for agricultural related information), market leading brands and those which are good at thought leadership stand out.
In a world with lower website click through rates, creating a reputation for putting out good reliable information will give you an advantage when it comes to searchers clicking AI links for more detail.
SEO is not dead, not in the slightest, but AEO is a new kid on the block. Like SEO is optimising for search engines, AEO is optimising for answer engines – things like ChatGPT and Google’s Gemini (which powers AI Mode).
Technical SEO has never been more important – from structured data to internal linking strategies and how any content served via JavaScript on a site is rendered, it’s hugely important for AEO. Content is still also hugely important and optimising it in a way that answer engines can extract data from.
A solid GA4, Google Search Console, and reporting platform/data analysis tool is key to understanding the impact of AI answers on your website traffic and performance.
You must be able to find out:
Using third party tools which allow you to discover keywords with and without AI overview panels can also be useful in the content planning process.
It’s easy to become complacent when something is working well and if your informational content is allowing you to report consistent organic traffic growth, sticking with what’s tried and tested is understandable.
However, AI overviews mainly impacts informational intent queries. Queries with more commercial intent are impacted less and users are far more likely to click through to a website if they’re looking to buy specific products or services. So, make sure you’re giving these keyword sets and sections of your website enough resource.
Also remember informational content not only drives direct search traffic but adds a lot of topical relevance to your website, in a lot of cases powering your key commercial pages and the rankings they hold.
If you want to survive the world of SEO, you need to keep calm and carry on. Don’t carry on regardless and in a blind way of course but do avoid panic and knee-jerk reactions.
SEO is dynamic and that’s why we love it. We find ways of delivering value through organic search regardless of the curveballs search engines like to throw at us.
Now where’s that bottle of whisky …
