Website Artificial Intelligence Optimisation (AIO): could AI ever replace search engines?
Jack Stanley ACIM
Digital Marketing Expert | Senior Digital Marketing Manager at Thistle Initiatives
The hype and pace at which AI is moving has accelerated changes in the way many of us consume information. Generative AI tools such as ChatGPT, and digital assistants like Siri (Apple) and Alexa (Amazon) are able to provide users with information at the drop of a hat, decreasing the need / desire for cross reference analysis and research.
Much like in evolution, technological advancement often leads to bad news for whatever came before, and this got me thinking whether Generative AI tools such as Chat GTP, are the natural successors to Search Engines like Google.
Probably unsurprisingly, the answer isn’t straight forward.
First and foremost, I think most digital marketing and SEO experts will agree that AI isn’t likely to challenge Google and its counterparts anytime soon. These are new technologies and whilst increasing in popularity, Statista estimates that less than 4% of the global population will use an AI tool this year (2024). In contrast, Google alone handles more searches a day than there are people on Earth (8.5bn).
However, the pattern of behaviour is undeniably shifting, and the desire for quick information, will almost certainly lead to an increased use in generative AI tools, for what we’ve all previously been using search engines for.
The question that follows for digital marketing and web experts is: is there such a thing as website AIO (artificial intelligence optimisation)? And how can I get the likes of Chat GTP to direct users to my website for relevant queries?
Here are my findings and thoughts below.
AI serves a different purpose to search engines
It’s important to note that generative AI and search engines aren’t as similar as they seem. When putting a query to Chat GTP, it will offer a response based on various sources and processes, whereas Google will provide a range of sources for users to formulate a response themselves, based on the information in the results provided.
Questions have been raised around the accuracy of what is being provided by AI tools, and most have made it clear that they may be prone to inaccuracies. Accuracy of responses should be a key consideration for AI developers to increase usage and challenge search engines, but the increased volume of misinformation on the web will be a key challenge. In the meantime, a more cultural question remains: will users value speed of response at the small risk of accuracy?
AI tools are ‘alive’
AI tools, not only continue to develop but the incredibly smart people that run them, but they also learn from its users. The demand for information will often pivot the sources and responses shared by the likes of Chat GTP. Satisfaction of a response and the queries that follow are key indicators to how these platforms will evolve.
For example, when asking Chat GTP: who are the best footballers in the world right now?, Jude Bellingham did not feature in the top 10. Whether anyone agrees with this is somewhat irrelevant, but if enough users ask the same question, and query why Jude does not appear in the list, Chat GTP will begin to re-evaluate its responses on this topic. In fact for this very scenario, Chat GPT changes its mind, citing “You're right[…]. Including him [Jude Bellingham] in the top 10 is indeed justified.”? See the full exchange here.
Now this isn’t a prompt for marketeers and web experts to spam Chat GPT with relevant queries, followed up by “what about my company?” in the hope of getting listed in all the right rankings… But the ability to challenge and query responses is key. High profile and high reach brand awareness marketing campaigns, could lead to long-term organic growth on this front, as these tools grow in usage.
领英推荐
Fact vs Opinion Queries
When querying facts, most AI tools will be able to associate a response based on standout and reliable sources. For example, when looking into financial regulation details, the FCA will be referenced. Or to continue the example around Jude Bellingham, when querying “how many goals did he score last season?”, Chat GTP will source its responses from (what it deems to be) official websites. The information provided is indisputable fact, and competing for these sorts of queries is somewhat futile, given the reliable and go-to nature of the sources used. In a similar way, Google users will often stop research after finding a reliable source with all the information they need.
However, assuming most readers aren’t the marketeers for the FCA or LaLiga, there is a real opportunity for parties in the relevant markets to lean into opinion-based-queries on AI platforms. For example, the responses to the query “why is Jude Bellingham so good at football?” will be diverse and subjective, outlining a plethora of reasons why the Real Madrid star is as good as he is, from a variety of sources. Some opinions may even think Jude Bellingham is not a good footballer, and articles to this nature may be referenced to provide a balanced response.
Seeking AIO opportunities based on topical and popular queries in your industry and knowing which of these you want your site to rank for is key.
How to optimise websites for AI
If you haven’t guessed where this is going, I went straight to the source for this one, and asked ChatGTP directly. The response will be music to SEO experts everywhere (even the AI sceptics) as optimisation techniques mirror widely used SEO strategies, citing:
Data and analysis
Unlike Google, ChatGTP does not provide much insight (publicly at least) into what it’s users are querying. Google has a variety of its own tools (Google Search Console and Google Trends) to help experts analyse their SEO performance, and there are many third party tools such as Semrush and Screaming Frog that will also provide insight on where efforts need to be made.
ChatGTP itself, quotes user privacy, data security and platform policy among a number of reasons why query data isn’t available. However, data is key to measure effectiveness of a website’s SEO performance and AIO will, in my opinion, only be a key consideration for web and marketing experts, once they are able to measure the outcomes of their efforts.
Conclusion
A lot has been made of whether AI will replace humans in a variety of capacities, but AI will first and foremost replace existing technologies. Whether Chat GPT will replace Google search remains to be seen, but as AI tools grow in number and usage, marketing and website experts will soon need to consider their approach to AIO (artificial intelligence optimisation).
Whilst AIO and SEO tread a very similar path for now, there is no guarantee that this continues into the future and marketing and web experts will need to evolve their strategies with the tools, in a similar way to how we adapt to changes in search engine algorithms.
Data availability and analytical advancement will be key factors in the uptake of AIO, with hurdles such as regulators being unable to match the pace of the innovation of AI, no doubt playing a part in wider use and transparency.