Search Generative Experience: What We Know So Far

With heavy hitters such as Microsoft, IBM, Meta and Google collectively reshaping digital spaces, business owners are wondering how best to navigate this new search environment.

I’ve talked about the relationship between SEO and artificial intelligence before, with the meteoric rise of language models such as ChatGPT and Bard reshaping the industry. Now, Google is preparing to shake things up with Search Generative Experience.

Dramatic, I know, but the significance of this technology should not be underestimated. If and when SGE is released from beta, it is set to transform the ways in which individuals discover, consume and interact with information online.

What is Google SGE?

Put simply, Search Generative Experience is a brand-new feature that uses advanced artificial intelligence to improve users’ Google search experience. SGE uses generative AI to provide concise overviews of search topics, offering users quick insights without the need to visit multiple web pages​ or scroll further down the search results. Answers to complex user searches are condensed to short, informative snippets, positioned at the top of the search results page. 

At present, a beta version of SGE is available in more than 127 countries and, while it has yet to launch in Europe, the tech giant’s self proclaimed ‘cutting-edge’ technology is certain to change the way that individuals discover and consume information. 

SGE’s UK release was predicted between January and March 2024 but, as of yet, there is no sign of rollout.

How does it work?

Google SGE is powered by a number of Large Language Model (LLM), including Pathways Language Model 2 (PaLM2), Multitask Unified Model (MUM) and now, Gemini. Each of these LLMs are capable of processing and analysing huge quantities of data, drawing relevant information from Google’s infinite knowledge base and using it to provide answers and create content. This is done via the implementation of technologies such as natural language processing, machine learning, and deep learning. 

We’ll discuss PaLM2 in further detail a little later in the article, but it is worth noting that this technology forms the basis of SGE and many Google Workspace applications, such as Docs and Gmail.

What’s All the fuss about?

As mentioned previously, a full rollout of SGE would mark a significant digital milestone, introducing artificial intelligence to the 92% of web users native to Google. In Google’s own words, they are ‘Supercharging Search with generative AI’. With that in mind, let’s explore the technology’s key features, capabilities and limitations.

Key Features

AI-generated summaries and topic overviews – SGE draws from a vast pool of information to provide concise summaries and overviews on a wide range of topics, enhancing user understanding at a glance. 

User-specific, customised search experiences – SGE tailors search results and recommendations based on the user’s previous searches, preferences, and behaviour, offering a personalised browsing experience.

Interactive search results and visual content – SGE presents search results in an engaging, interactive format, incorporating rich visual content such as images and videos to complement textual information. Specifically, SGE is ‘designed to help people more deeply engage with long-form content from publishers and creators’ and make it easier to learn while you search.

Prompt-led navigation buttons – SGE allows users to continue their search journey through the use of intuitive prompts, facilitating easier exploration of search results and related topics without the need for manual input. 

Comprehensive search results, incorporating multiple content types – SGE delivers extensive search outcomes by including a variety of content types like articles, videos, and images, ensuring a thorough exploration of topics.

Limitations

Misinterpretation – No language model is perfect, and Google SGE may misunderstand the context or intent behind a user’s query, leading to irrelevant or off-target search results.

Hallucinations – AI is prone to generating incorrect or fabricated information within its responses, presenting challenges in ensuring the accuracy and reliability of the content provided. This imagined information is known as hallucination.

Biassed Answers – In addition to misinterpretation and hallucination, there is also the potential for Google SGE to exhibit bias in its responses, influenced by the data it has been trained on. Google commented on this, stating that ‘despite additional safety and bias prevention guardrails applied, some nuances of data patterns still seep through, and can lead to SGE producing biassed results.’ That said, it is worth noting that in some instances, attempts to ‘un-bias’ language models have backfired, leading to skewed representations and offensive depictions of gender and race.

Trust – Lastly, an amalgamation of the above mentioned points means that many users do not trust AI technology to produce reliable results. This is absolutely crucial as the long term success and development of SGE depends on user acceptance.

How will SGE affect your business?

Google is no stranger to algorithm updates, regularly rolling out unannounced changes that impact search ranking and leave business owners scratching their heads. That said, these minor hiccups are relatively inconsequential in comparison to SGE, which may well be the most significant change to Google search, and eCommerce, since its inception. 

The introduction of a new SGE results section sees organic, paid and sponsored entries pushed down page one. In other words, SGE takes up a lot of SERP space, making it the first thing that users see when they search and, consequently, valuable organic real estate.

With these changes, SGE is reshaping user journeys, stepping away from traditional methods of search and towards a much more conversational experience. In doing so, Google will massively reduce the likelihood of individuals clicking through to specific web pages in order to obtain the information they seek. Instead, personalised, hyper-relevant results will be provided via short form snippets, images and artificially generated responses right below the search bar.

Data from a recent study demonstrates this shift away from traditional search ranking, and towards a much more dynamic methodology. Following a basic search request, an alarming 93.8% of the generative links came from sources outside of the current top-ranking organic domains. 

For businesses, especially those operating within the eCommerce sector, this transformation is simultaneously problematic and promising, begging the question, how does SGE source its information? 

Much to the frustration of business owners and digital marketers alike, at this point in time there is no sure answer. The language model used to power SGE, known as PaLM 2, is closed, meaning that, though it may be trained on publicly available data, Google’s AI model is not publicly transparent. For this reason, SGE’s specific data collection and extraction methods are topics of frequent debate among digital marketers and search specialists.

How your business should be adapting to SGE

For online businesses, optimising your digital presence in anticipation of Google SGE is absolutely essential; not only will this technology reshape the search engine results page, it will transform the ways in which users discover and interact with your website and products. 

As mentioned above, the specific means through which Google SGE collects and extracts data from sites is unclear. What we do know, however, is that the movement towards a conversational search experience places further emphasis on relevant information. Businesses should maintain a sharp focus on strategy and content, creating authoritative, useful pieces that align with Google’s E-E-A-T requirements.

Additionally, businesses looking to optimise their sites for SGE should focus on proper auditing and implementation of schema structured data and meta information. Everything from product descriptions, prices and reviews to page types and meta descriptions should be up to date and properly structured in order to maximise visibility within SGE’s framework.

It is important to understand that, significant though it may be, SGE is representative of a much larger shift; artificial intelligence has, and will, transform the world of digital marketing in more ways than one. Maintaining online visibility during such turbulence requires a clear cut strategy, consistent, high-quality content and a multi-channel approach. 

Here at Superb Digital, we’ve been monitoring the development of SGE for some time and are already working on ways to optimise site content for AI-generated search. Our team is keeping a close eye on developments and, when SGE launches, we’ll be here to provide guidance on how to best leverage this technology in your digital marketing.

To discover how we keep our clients ahead of the game in an ever-changing digital landscape, book a discovery call today.