What is AEO? AI Engine Optimisation Explained
As the world continues to rely more on AI chatbots and LLMs, businesses will need to learn how to operate in a market where SEO matters a lot less.
If you ask BING about Europe's first Generative AI consultancy, the Chat-GPT powered bot will bring you straight to ACQUAINTED. Sadly, on the SEO-side of things, we don’t do as well — Acquainted also happens to be a song by The Weeknd and, with almost 15 million views, is proving to be one particularly slick problem for us to hurdle. Fortunately for us however, every day this becomes less of an issue.
From SEO to AEO
That is because SEO is changing. For years, SEO has been the driving force behind the internet as we know it, with companies either incorporating white hat practices to come top of search results or just flat out paying for that privilege.
But since the world went GenAI doolally, SEO is having a bit of an existential moment. The problem that LLMs pose for SEO is clear enough for all to see: by giving direct answers to queries rather than simply pointing users to suggested links, LLMs like ChatGPT create a world in which there will be much fewer searches, uprooting the basis for SEO and the value of all of that search optimised moolah to boot.
It’s not only us who think so. In May, Bill Gates stated that Generative AI could end up meaning that users never visiting a website again. Back then, we were calling it Chatbot Optimisation, but after a few months of maturation we are beginning to see the key characteristics (as well as a new name) to this new approach.
AI-Engine Optimisation (AEO) is how businesses will have to maximise online content for visibility and discoverability in the age of Generative AI. In general, AEO involves directly optimising for AI, rather than relying on traditional SEO methods. As personal AI assistants like ChatGPT become more prevalent, traditional SEO methods that rely on crawlers are becoming less effective.
From APIs to Plugins
One key aspect of AEO is feeding real-time data to foundational AI models via APIs (Application Programming Interfaces). This can help improve the accuracy and relevance of their responses, making it more likely that your business will be recommended to users.
A quick recap on APIs for the uninitiated: APIs allow different software programs to communicate with each other and share data. In the case of AI assistants, APIs allow the assistant to access live data from many different sources — such as websites, apps, calendars.
For example, if you run an e-commerce website, you could provide real-time information about your products, prices, and availability to AI models via an API. This would allow the AI model to have up-to-date information it can provide to users who are searching for products similar to yours. By connecting to a database, the API would allow the assistant to see up-to-the-second information on what products are in stock, current prices, product details like size and colour (and more).
We are already beginning to see this method develop in the wake of ChatGPT’s plugins. Plugins are tools designed specifically for language models that help ChatGPT access up-to-date information, run computations, or use third-party services. These APIs are aligned with the business that creates them, and will prove to be a key aspect of AEO.
By developing plugins, enterprises can help improve the accuracy and relevance of AI recommendations, making it more likely that their business will be discovered by end users. For example, if someone was to run a recruitment company, they could develop a plugin that connects those searching for a job to their database. In this scenario, potential candidates need only explain their background to the model itself, and they will be shown the jobs that suit them best.
Branded LLMs
The step up from developing a plugin is of course for a company to build their own large language models (LLMs). These models will replace their customer service bots and will be based atop their own datasets completely. Apart from the security and IP-safe bonuses of this walled garden strategy, having your own LLM can help improve customer engagement and satisfaction by providing more accurate and personalised responses to customer inquiries.
On top of that, not many, if any at all, brands will want their user experience to be to the tune of ChatGPT. Instead, to ensure a more branded user experience, these LLMs will be trained on an organisation’s own tone of voice, with appropriate levels of empathy and emotional intelligence built-in for the product and situation.
This in itself is a sea-change from how SEO worked before. In AEO, it will no longer be only about stating the facts, instead, it will also be about blurring the line between technology and human interaction, by incorporating a design that incorporates how the end user will also feel.
An Omni-Channelled Approach
Whilst AEO is still in its nascent stage, at the end of the day, it will prove important for businesses to adopt a dual approach that blends SEO and AEO in an omni-channel strategy. Let’s be real here, it is unlikely that Alphabet (Google’s parent company) is going to be pulling the plug on SEO — which accounts for two-thirds of its income — any time soon.
But that is not to say that delving into more modern and complex solutions is not worthwhile for businesses looking to gain competitive advantage. At the end of the day, those who become pros in both SEO and AEO will be best placed for the years ahead.
But for now anyway, a majority of LLMs still crawl the internet for data. Counterintuitively then, for companies who wish to be front and centre of these models, a a better strategy for now might be to let these LLM’s hoover as much data about your company as possible.
In essence this means that the tried and tested SEO method of having better, higher quality content is the best way to get your business into these models. Until it’s not that is.
If you would like to learn about how to prepare your organisation for an internet driven by AEO, get in touch today.