Google Search Generative Experience 101

The future is here – Google's new Search Generative Experience is about to reinvent search as we know it. And the update seems to be just around the corner.

Announced at Google's I/O developer conference on May 10, 2023, SGE aims to move beyond the traditional ten blue links. Instead, it will directly provide conversational, contextual answers on the search results page.

While still in beta testing, it’s already clear that SGE can drastically change the search game.

This article offers a sneak peek at what we know so far about Google's AI-powered search. I'll check out the new-look results pages and tackle key questions about how SGE could change the search experience and SEO.

So, let’s dive in!

What is Google SGE?

Google announced Search Generative Experience to integrate AI more deeply into search. SGE utilizes generative AI to provide enhanced, conversational results directly on Google's search engine results pages.

Up until now, Google search has relied on crawlers indexing websites and algorithms ranking content. SGE adds to this by using a large language model called PaLM 2. This AI model is trained on a ton of data to infer contextual answers to queries, instead of just pulling up indexed pages.

The goal of SGE is to transform search from retrieving a list of blue links to having a dialogue with users. Traditional Google search still works the same, but SGE adds AI-generated responses to make the results more relevant and intuitive. Here is an example:

SGE marks a major shift for Google search through the integration of generative AI. While still in beta testing, it provides a glimpse into the future of search powered by models like PaLM 2. 

Once the new features go live, Google is expected to become conversational, going beyond keywords to understand full contexts and questions.

The early version of Google's SGE: a quick overview

Search Generative Experience is still in limited preview mode. Up until recently, it was only accessible to a limited number of users in the US, India, and Japan, and it only worked in English, Hindi, and Japanese. 

But, as I was writing this article, Google introduced a new update announcing that SGE is expanding the experiment to 120 more countries. Plus, four more languages are now available: Spanish, Portuguese, Korean, and Indonesian. 

Besides that, here’s what I’ve noticed during my hands-on experience with Google’s SGE:

How SGE changes the search experience

In the past, you'd have to ask Google multiple specific questions to get complete information on a topic, like:

But with SGE, you can now ask one broader, conversational question and immediately get a comprehensive answer.

For example, you could simply ask "What DSLR camera setup should I get as a beginner interested in portrait photography?" Rather than just showing relevant webpages, SGE can generate a custom response synthesizing details from multiple sources to fully address what you're asking and why.

Here’s another example straight from Google. A complex question is easily answered by the new AI.

The goal is for SGE to handle complex informational needs in one search instead of making you rephrase multiple queries. It aims to directly give full answers to open-ended questions through AI summarization and recommendations.

Accessing SGE through Search Labs

Google introduced the AI-powered search prototype in its new Search Labs program.  After the recent update, Search Labs is now available to people in 120 more countries.

Search Labs allows select users to test early-stage search projects before wider release. It launched in May 2023 for US testers interested in shaping the future of Google Search.

Users can try out new and creative ways to search and explore info in Chrome on desktop or the Google App on iOS and Android mobile devices. If you're accepted from the waitlist and click the Search Labs icon, you'll see experiments you can toggle on or off.

To try SGE yourself, you’ll need:

By joining Search Labs and turning on the SGE experiment, you get to be one of the first to preview Google's vision for the future of AI-powered search. And you can directly influence how the tech evolves by giving your thoughts on what works and what doesn't.

When will the Google SGE go live?

Google has not announced an official rollout date for Search Generative Experience. The feature is still in the experimental phase. 

According to insights from Google's Search Labs team, there's a possibility that the SGE experiment might wrap up by December 2023, as suggested by Barry Schwartz.

How SGE changes Google's SERP

SGE prototype totally reimagines the look and function of search results pages. Even though it's still in early testing, SGE gives a sneak peek at how AI could transform Google Search.

The most obvious change is the interactive AI-generated snippet at the top of the results. This condenses and summarizes information to answer the user's query, with supporting evidence from websites in a sidebar carousel. The snippet evolves in real-time as the AI processes the request. Here’s what it looks like:

Unlike current results, SGE snippets are more conversational and detailed. They allow for natural follow-up questions without needing new searches. The snippets adapt on the fly, continuing the dialogue while citing sources.

The AI responses can take different visual forms depending on the query type:

No matter the format, the AI snippet and website carousel take up a lot of space on the screen. This pushes traditional text links lower on the page, especially on mobile. The emphasis shifts away from blue links to conversing directly with the AI assistant.

Now, let's take a closer look at the main types of AI snippets:

Types of AI snippets in SGE

General Informational Queries

For general informational queries, SGE generates a conversational snippet that summarizes the answer to the question. It includes 3-5 links to websites used as sources, with quotes and images pulled from those sites. 

Follow-up questions are suggested to allow further discussion with the AI without new searches. How-to queries see similar results in a list format. The goal is a detailed back-and-forth exchange to provide comprehensive information.

Listicle-Style Informational Queries

Queries best suited for listicle-style responses prompt SGE to generate a bullet list of relevant solutions or steps. Each list item has a short description. Sometimes this is combined with a paragraph summary above the list. 

List snippets may also integrate maps for location-based queries. The AI synthesizes advice in an easy-to-read listicle based on user intent.

Local Searches

For queries with clear local intent, SGE produces a local pack-like display. It includes up to five relevant local businesses instead of the traditional three. 

Listings contain typical local pack information like photos, maps, reviews, contact details, etc. The AI pulls the most useful local business information to answer location-based searches.

E-commerce Queries

Product-related keywords lead SGE to generate shopping-focused snippets. They recommend products they think are suitable for the user, along with descriptions, images, pricing, seller info, and user reviews.

The format resembles a shopping carousel. SGE taps into Merchant Center data to get more product details. The goal is to act like a product finder/recommender based on what the user searched for.

When will Google’s SGE not generate AI snippets?

There are certain types of searches where you likely won't see an AI-powered response. These include:

So navigational terms, pre-formatted quick answers, sensitive subjects, and queries where expert guidance is preferred generally do not produce AI conversational snippets. SGE seems more likely to respond to informational, product-related, and local searches. But policies on appropriateness continue to evolve.

Examining when Google's SGE search engine does or does not generate results can provide insights into the ethical safeguards put in place by Google. There seem to be certain topics and questions where SGE defaults to withholding a direct answer, instead recommending the user seek answers elsewhere. 

This reveals that Google has prioritized human judgment and discretion over pure algorithmic responses for sensitive or controversial subjects. Further testing could continue to map out these algorithmic limits.

Google's SGE search prototype displays different messages when AI-generated snippets are not provided. Let’s briefly go through them:

What will happen to existing SERP features, such as featured snippets?

Google's new AI chatbot responses have raised questions about the future of existing SERP features like knowledge graphs, rich snippets, and so on. 

For now, featured snippets and other common SERP elements are still showing up, even alongside the AI responses.

From what we know, the featured snippets appear right below the chatbot text, which is interesting since both aim to quickly summarize information about the user's query. Other familiar parts of the SERP, like "People Also Ask" boxes, videos, rich snippets, knowledge graphs, and local packs are also still around for now.

However, some SEOs predict Google may eventually remove featured snippets, local packs, and potentially other redundant SERP features in the future. The thinking is that the AI responses make those other elements unnecessary since they take up space to show info the AI responses already cover. 

However, Google has not confirmed if or when SERP features might be streamlined to focus on the new AI component. At the moment, the chatbot responses seem to complement existing page features. We'll have to see how Google balances the new AI capabilities with traditional SERP elements as it evolves its search tech.

Bugs and concerns found so far

Google's new AI-powered search has garnered significant attention and sparked some concerns within the SEO community and among website owners.

While this technology is still undergoing testing, a few key issues have cropped up that deserve attention:

Google has acknowledged the feedback and experimental nature of SGE. The company says it is working to improve attribution, relevance, accuracy, and handling of sensitive queries. But for now, the system has clear limitations that website owners and SEOs are documenting in hopes of future enhancement. It remains to be seen how SGE will evolve over time as Google trains the AI models powering it.

SGE and SEO: what to expect?

The websites linked and cited in SGE's AI-generated snippets often do not coincide with top-ranking search results. Apart from that, here are some additional observations from users so far:

So while SGE doesn't simply favor top-ranked sites, it does look for sources covering the specific nuances its AI responses emphasize.

By reverse engineering which websites get linked for different queries, publishers can better tailor content to how SGE synthesizes information. Matching the AI's unique angles and interests seems key to getting your site included.

Besides, industry experts have shared some more insights on the impact of generative AI on SEO and its implications with the advent of SGE.

It's anticipated that Google's AI will favor comprehensive, authoritative content, increasing the likelihood of such content being featured prominently in AI-generated summaries. This is particularly true for content that meets the standards of Experience, Expertise, Authoritativeness, and Trustworthiness (E-E-A-T).

There's also a projection that on mobile devices, AI-generated responses may dominate the screen, potentially leading to a decrease in organic traffic for classic links that appear below the new AI window.

To adapt, content creators might need to focus more on conversational, long-tail keywords to enhance their content's visibility in these AI responses.

Sites with in-depth coverage of specific topics might find themselves at an advantage. Google's AI tends to prefer detailed, context-rich information, which means sites that delve deeply into particular subjects could be more appealing for the search engine's AI-generated summaries.

However, there's an expectation that informational content could see a decline in traffic. With AI-generated responses providing direct answers on search engine results pages, users may not need to click through to the actual websites as often. This trend might be less pronounced for complex informational queries that benefit from more detailed, experience-based content.

Personalized or not? How SGE snippets work

Since its launch, some have wondered if SGE skews results based on personal data like standard Google Search. However, Google insists SGE currently operates ad and history-free.

According to official comments:

So, for now, Google states SGE focuses only on individual queries without broader tracking or profiles influencing results. As the feature evolves, native ads will likely be blended in over time. However the initial rollouts aim to provide impartial, context-based responses, according to Google.

How SGE could affect website traffic

Google's new AI-powered SGE could soon shake up search results as we know them. SGE lets users summon AI-written overviews to summarize search queries, diverting clicks away from regular organic listings.

Early projections estimate the technology could potentially divert up to 30% of organic clicks away from standard listings, similar to the impact of featured snippets. However, real-world usage patterns will determine the actual effects once SGE is launched.

What we can expect: Google's SGE feature has the potential to significantly impact organic search results and traffic flows.

Human behavior will be key here. If generating summaries is time-consuming, impatient searchers may not engage with this feature often. Also noteworthy is where users focus their attention - the left side of overviews where summaries appear or the right side with cited sources.

Early birds in launch markets will shape SGE's trajectory. A delayed EU rollout could allow more preparation time before the impact is felt there. Some see e-commerce as poised for gains, though, if product images in AI overviews attract high engagement.

The extent of SGE's impact remains unclear for now. But one thing's certain - SEOs will need agility and creativity to prepare for the changes ahead.

Which industries/types of websites could profit from the SGE rollout?

It's still too early to fully predict how Google's Search Generative Experience will impact SEO. The experiment is new and evolving quickly. However, some potential effects are emerging for different business types:

The impact on organic traffic for all sites remains to be seen. Local businesses and retailers may benefit most from AI, while informational sites could lose search visibility. But it's still early, and Google's algorithms will likely evolve. SEOs will be tracking SGE closely to adapt strategies as patterns emerge.

Best ways to brace yourself for SGE

There's been a lot of buzz about Google's SGE and its potential influence on SEO.

SGE is a work in progress, and its impact is still up in the air. Nonetheless, there are actionable steps you can implement to get your website and content ready for what's to come:

The bottom line is don't panic about SGE yet. Stay focused on robust SEO strategies that deliver results now while keeping an eye on SGE developments. As we learn more about how SGE impacts search, you'll be in a strong position to adapt and optimize accordingly. Patience and vigilance are key.

Wrapping up

Google's new AI-powered Search Generative Experience is poised to revolutionize search as we know it. While it's still in beta testing, SGE marks a significant shift in how we interact with search engines.

However, we should remember that the full impact of SGE remains uncertain. SEO professionals and website owners need to stay vigilant, focusing on tried-and-true SEO strategies while keeping a close eye on developments.

In a nutshell, SGE is the future of search, and it's evolving fast. So, stay in the loop and be ready to roll with the changes to make the most of this shift.