AEO, GEO, SEO and the Future Of Search – Some thoughts in August 2026
To say there is a lot going on in search right now is an understatement. I have been in search for nearly 15 years, and outside of big algorithm updates like Panda and Penguin (and even those are tiny compared to this), the launch of AI is by far the biggest thing to happen in SEO.
What Does AI Mean for Search?
Outside of the noise and the mad race every SEO is running to say they have “cracked AEO” what does it really mean for searchers, businesses, and SEO agencies like 427?
From an agency point of view, we obviously have a different search landscape, but on the other we still need to keep our eye on the daily work to deliver results for the millions of people who are still searching using “traditional” search. It is critical we balance the race to AEO and the real-world delivery of leads. Yes, lots of people are using LLMs for search, but those numbers are currently still dwarfed by the number of people using Google Maps and normal search.
We also have to remember that LLMs use traditional search for grounding and information on most searches, so Google and Bing still matter. Even Google stated that good AEO is simply good SEO. How long this stays the same, we do not know.
We are deep in research, testing, and monitoring, but we are also still deep in day-to-day delivery. We are really keen to avoid the race to “sell” the answers for more and more money. It is very clear some people are doing just that and charging a lot of extra money for AEO when most people in the industry know full well the data isn’t there to really prove the work they are offering does anything.
Pragmatism, research, testing, and results are key for us. This will allow us to deliver what we know works and avoid burning client budgets on guesses and ideas that have no data to back them up.
How and Why People Use LLMs for search?
It is really interesting to take a moment to think about how we use LLMs when searching. What we are seeing and doing ourselves is actually using these models for things we may not have ever gone to Google for. Car repairs are one example. None of the team are mechanics and would normally not risk getting involved. However, we are now turning to LLMs with pictures of the engine bay and questions about what things are and why they are making a funny noise. The results often sound very impressive, and it’s amazing how much these LLMs look like they know. Would I have done that with Google? No. So we are bringing search into areas we would not have done before. Rather than replacing traditional search, it is augmenting it. We then tend ot go back to Google to order a part or find a garage rather than take that search all the way through with the LLM. This is because the answers are too long and take a while; we also see some of the links don’t work, and the choices for shops, etc are not always great.
So rather than immediately replacing search, LLMs are taking more of a knowledge role, a deep search offering that then leads us back to a quick Google search to spend money. This is an important point; the buying and real lead generation still comes from Google. The speculative information stage is coming from LLMs. The mad rush to get cited in LLMs above all else may be a mistake…for now, because the conversions are often coming from traditional search (often, not all the time).
Once again, the key here is balance. We are very much gunning for LLM search, AEO, GEO….whatever you choose to call it. But we need to keep in mind the commercial opportunities for businesses in search as well.
What is an AI or LLM citation?
People talk so much about getting cited in LLMs, but a lot of people don’t seem to fully understand what that means.
A citation in a AI engine can be one of 3 things
- The information on your website is being used to create an answer in an LLM
- Your brand is mentioned as a potential resource or suggested business related to the prompt
- Your website is listed and linked to in an answer.
Only 2 of these are of any use to a business. Most of the time, a citation will be the first one on the list, and so the information is being used to form a result with no reference to the business at all. It is really important to keep this in mind when looking at citation data and celebrating gains. You may have been cited 1000 times, but unless there is a brand mention or a link, there is very little value. You could, and should, argue that being picked up at all is a good sign and it may lead to more detailed mentions later, but the context here is important.
What is AEO and is it Different from SEO?
For 427 Marketing and a lot of SEO specialists, AEO is simply SEO with extra tactics and tasks. What we do is optimise websites to drive more traffic and leads from search engines. An LLM like Claude is just another search engine so SEO naturally includes work to improve performance there as it does in Google or even YouTube. However, that is a slightly facetious way of looking at it.
AEO is a growing set of tactics that we can deploy on and off a website to improve the number of times a site is used by LLMs for form results or to ground their learning when creating results.
SEO is a very large set of tactics on and off a website that help to improve the rankings, impressions, traffic and leads generated from people using search engines that include Google, Bing, and many others.
These two things are intrinsically linked and should really be treated as one process. All the above tactics have beneficial impacts on all related search engines so the end result of the collective work should be gains across all platforms. Ergo; SEO is AEO and AEO is SEO.
AEO is SEO, but with extra new tactics we need to deploy for specific gains.
Is there a growing negative attitude towards AI and LLMs?
For most of us, the rise in use of AI in the workplace is massive. Search is growing too, but it is day-to-day work where we are really seeing the growth. But there are some rumblings of issues and some change in attitude even now.
The Environmental Impact of AI use
It is safe to say most people know there is an environmental cost to using AI. The quotes about the amount of water each use takes are numerous, the new data centres being built, the energy required, etc. As I write this in the middle of a heat wave, these things are going to matter. There may well be a pushback on AI use coming.
One example may be B Corp companies. These companies have gone through rigorous assessments to show they are acting in a sustainable manner. But did this assessment include AI use? Unlikely. What happens to a business that adopts AI use in a massive way but also has B Corp status? Will this have an impact? And should it?
AI use in business does have an environmental impact, but how big that impact is, we do not know. We may well see consumers making choices based on this in the future. We may need to start stating what we do as a business in terms of AI use to allow consumers to make an informed choice on sustainable practices and our eco credentials.
The impact of AI on the creative industries
People who work in the creative industries, including video, marketing, design, music and art, are all feeling the pressure of fast AI-produced work that some people are choosing to use. This is already starting to cause pushback. While it seems quick and easy, we are all noticing how “AI” every poster we see looks, and there is already a growing disdain for this kind of thing. A time will come very soon when things have to be labelled as AI-generated, and it will be interesting to see how consumers react to a business who sues AI to produce an advert rather than real people. Younger generations are already calling out AI on social media because they are getting fed up with it. AI-generated music is another area where people may start to make very clear consumer choices to avoid any association with it.
This is not just because it can be bland and feel fake; it is also because it involves supporting people’s jobs and livelihoods in the creative sector.
Will we see “Peak slop” and then suddenly businesses will start paying a high price to get real creatives back on board so they can confidently claim there is no AI used in their new TV ad or social media videos?
Opening the black box on AI and LLMs
Let’s get back to AEO and LLM search. The reality now is that we and many other agencies are working on a wide range of tactics to make sure we deliver good results for our clients from the growing use of LLMs as search engines. But until very recently, the data about what is actually happening has been very thin on the ground. There are lots of AI trackers out there, but when you get into the nuts and bolts of it they are simply not covering enough ground to give a reliable result. By these, I mean they are running maybe 5 prompts per search term per day per LLM and then reporting on how many times a site was mentioned. The problem here is people will use a vast array of prompts, including a vast array of terms. Narrowing these down to 5 or even 15 is still far too narrow. Each LLM uses different weighting for each letter, let alone full words, and some will be using previous searches to make choices as well as the prompt. Add to that the fact they are updating all the time, and you end up with data that is so far away from being statistically significant as to be largely pointless. Yet we are seeing some very high costs for software offering this kind of data.
The back box needs to open from the inside and not from the outside. By this I mean we need to see data from the specific LLM platforms about grounding queries and citation rates, and not software guesses from external companies. This is now starting to happen with Bing/Clarity, and we will see more and more of this in the coming months. We will have Search Console-style data for most of the LLMs, and when we do, we will finally be able to see what AEO is really about, and I predict a number of tactics people have been claiming to use will be shown to be pointless. The data we are now seeing from Bing is very interesting indeed. We can now see how often a site has been used to ground an LLM prompt. So essentially we don’t know what the prompts are, but we can see a specific site has been used by the LLM to check its facts when looking for something like “SEO agency in Sussex”. The prompt could have been very complex, but we get a number like 30%, which shows us the site was used in 30% of all prompts related to that query. This means we can start to work on those terms, and now we can measure something we can improve it! We are also getting citation data through, so we can now map out gains based on the work we are doing.
At the time of writing this, this is big news. It won’t be in 2 months and may seem very out of date in a year…sorry!
What does this all mean for business, agencies and users?
It means everything and nothing. Good SEO now must include AEO. But good AEO is also good SEO so for people working to improve rankings, the job just got bigger, but it always has. SEO is an ever-changing world, and this is no different.
For users, it is simply a matter of having more choice. Sometimes AI will be suitable; other times it will give bad results, cause harm to the environment, produce bad art, or just not be the right thing. Either way, the choice is there.
For business, this may mean a bigger budget on SEO because the amount of work needed is growing. We can also throw in things like video and social media, now falling into the AEO/SEO world too so that means more alignment between channels. Everything is now SEO, and SEO is now everything.
