Are AI Overview Changes a Step in the Right Direction?

After a controversial start, Google is testing a new approach to search-integrated AI Overviews - which may come as welcome news to web publishers.

Currently rolling out more widely after initial testing in the US, the new format looks to address criticism by including more prominent ‘source’ attribution in the form of ‘inner links’ within the AI ‘answer’ itself, supported by references and some search results in the right-hand panel.

In a positive turn of events, this new implementation should make it easier for sites to occupy ranking positions at the very top of the search once again. While AI Overviews are still far from perfect - larger publishers seem to have an effective monopoly over surfacing in this way - it is arguably a step in the right direction, both from an ethical point of view and a commercial one.

Why is a lack of source attribution a problem?

One of the biggest concerns around AI-powered search and large language models like ChatGPT more generally is the lack of source attribution, which makes it very difficult to verify the information presented and assess its credibility.

Search engines are incredibly powerful sources of knowledge because they give us access to vast amounts of information from countless sources. However, it’s ultimately up to us, the users, to interpret that information and draw our own conclusions from the searches we conduct.

Yes, the search landscape is far from neutral and Google’s influence on search engine result pages (SERPs) gets stronger and stronger with every core update - but we still have a multitude of sites and sources to analyse when we want to make our minds up about something.

What happens if people stop making their own decisions on what is a credible source and what isn’t?

People aren’t necessarily lazy, but they naturally use mental shortcuts to make decisions when faced with information overload and uncertainty. 

According to the principle of parsimony, the simplest/most economical explanation is often the most likely to be correct. It’s also referred to as Occam's razor, as we “shave away” unnecessary information.

That being said, can an experimental algorithm really make that decision for us? Can it take away our agency in deciding what’s true and what isn’t?

Liz Reid, VP and Head of Google Search, earlier this year published an article titled “Generative AI in Search: Let Google do the searching for you”. 

While this feature cannot be turned off completely, there is currently a small way around it: by adding “-ai” to the end of your query, you can disable AI overviews for that specific search.

How are things changing for SEO?

On top of the ethical discourse, Google’s initial choice to not include any sources or links in their “AI Overviews” initially caused a lot of raised eyebrows in the SEO industry.

With the first result page full of AI overviews, featured snippets and sponsored PPC links, the organic battleground has arguably become more challenging as ever before. It’s understandable that publishers may feel the “helpful content” that Google asks them to write is now being lifted to power unattributed AI Overviews, rather than delivering hard-earned traffic.

For their part, Google has argued that AI overviews help increase websites’ traffic quality and the proportion of qualified users they receive.

Regardless, with this latest update, it looks like steps are being taken to draw attention back to the actual content, rather than just the ‘AI version’ of it.

How does ICS stay on top of the ever-changing search landscape?

Especially in the last couple of years, the SEO landscape is known to be volatile and sometimes unpredictable, due to new technologies, algorithm updates and competition shaking up the status quo.

At ICS-digital, our commitment is to deliver value in the form of rankings and traffic for our clients. In order to do so, our SEO experts regularly adapt our SEO and content strategies to make sure they stay relevant and effective no matter the many Google updates. This might include executing in line with published best practices, while in other cases we may rely on testing and experimentation to validate what works well - and what doesn’t - when it comes to organic search.

Get in touch to find out more about how we can optimise your on-site strategy for the world of AI-powered search.