A Big Shift in SEO: What Google AI (SGE) Means for Website Traffic
On the 10th of May 2023, Google announced its plans to completely change search by adding AI-generated answers.
In anticipation, SEO enthusiasts got scared — claiming this is, yet again, the end of SEO.
Most niche sites will die if Google and other search engines continuously reply to informational queries with AI.
— Hasan Morshed (@hasanmorshed_) February 16, 2023
Admittedly, seeing Microsoft integrate generative AI into Bing a few months ago worried me about how Google would respond. So like many, I watched fellow Aussie Cathy Edwards, the VP of Engineering at Google, demo the future of search at Google I/O. Here’s the video.
To summarise, you get your question answered by AI, reducing the need to click through to websites.
What does this AI search engine mean for building organic traffic to websites?
In this article, I’ll dig into the interesting sources I found about this upcoming change and share what you can do to continue driving organic traffic to your website.
In addition to the video above, I’ve reviewed Google’s official PDF, their article on generative AI, tech publications’ take on Google AI, endless YouTube videos, and Brandon Gaillie’s podcast on traffic forecasts for publishers (more on this later).
To get started, let’s understand what Search Generative Experience means.
What is Search Generative Experience?
Search Generative Experience or SGE, is a term you’ll be hearing a lot in this new iteration of Google Search.
With SGE, you can ask new types of questions in Google to get an AI-generated response supported by links to more resources. You can also follow up your initial search with more questions to refine your answer. It’s basically ChatGPT in Google Search.
This interface of this new search experience is what Google calls AI-powered snapshots.
Until now, Google Search has trained us to type specific queries to get results. It’s hardly natural. We literally communicate like a robot. Instead of asking “How do I fix a leaking faucet” we write “fix leaky faucet”.
SGE turns rigid search queries and keywords into a conversation — opening up a new world of questions and data.
SGE is a necessary step in the evolution of Google. It provides a richer experience as the AI is your web assistant to source the most relevant information for your queries.
But by front-loading so much information, what does this mean for getting website traffic?
How Google’s AI-Powered Snapshot Will Impact Website Traffic
In Google’s demo, they used an example of a search query, “good bike for a 5 mile commute with hills”.
You get the answer front and centre in the search result (above in blue). This is helpful as you get the following:
- A neat summary of features to consider for your bike
- Bike recommendations
- Link to three articles for further review
The three articles represent the new top results in Google, with the rest of the links pushed below the snapshot. But if you’re satisfied with the AI’s response, will you click on the articles?
To forecast the traffic impact of the AI snapshot, I’m referring to an analysis by Brandon Gaille, a highly successful blogger and the founder of the SEO tool RankIQ.
Using the data from SEMRush featured snippet study together with when SGE will be rolled out in the US, here are the potential organic traffic losses categories by industry:
| Industry | Traffic Loss Percentage |
| Travel | 21% |
| Computers & Electronics | 21% |
| Arts & Entertainment | 21% |
| Science | 19% |
| Finance | 16% |
| Pets | 16% |
| Sports | 16% |
| Business | 15% |
| Jobs & Education | 15% |
| Food & Drink | 14% |
| Beauty & Fitness | 14% |
| Health | 14% |
| Hobbies | 12% |
| Books & Literature | 11% |
| Home & Garden | 10% |
Brandon notes that the traffic drop will likely fall from 10 to 21% for the industries not listed above.
While these are just estimates, it’s pretty alarming. At the same time, it’s empowering to have an idea of what to expect so you can plan.
Featured snippets shook up the SEO industry in 2014. Back then, we had the same concerns as we do now with AI. But like featured snippets, not all hope is lost for publishers regarding SGE.
4 Reasons Why Google Will Continue Sending Traffic To Websites
The AI snapshot is dynamic. For some queries, it shows website links front and centre — like this restaurant recommendation:
This is a positive sign for publishers. But beyond these types of results, there are four reasons Google needs to keep sending traffic to websites.
Reason #1 – AI Needs Information Sources
The development and training of AI systems heavily rely on data and information sources, particularly for natural language processing and understanding tasks. Websites serve as vital information sources that enable AI algorithms to learn and improve their understanding of various topics and user intents.
So if websites stop producing content because they’re not getting any traffic, what will the AI learn from? It’ll be a pretty incompetent AI if it can’t train itself anymore.
Reason #2 – People Prefer Reading From People
Google knows that people prefer learning from other humans. Don’t take it from me. Here’s a quote from Cathy Edward’s Google I/O presentation:
“Search will continue to be your jumping-off point to what makes the web so special. Its diverse range of content from publishers, to creators, to businesses, and even people like you and me. You can check out recommendations from experts like the National Parks Service and learn from authentic first-hand experiences like the Mom Trotter Blog. Because even in a world where AI can provide insights, we know that people will always value the input of other people, and a thriving web is essential to that.”
Google reminds us three more times of this point in their release documentation. Like when they know that SGE users will quickly get the lay of the land on the topic with links to relevant results to explore.
Or that AI snapshots are a jumping-off point from when people can explore a wide range of content and perspectives on the web.
Or that people want insights from others to inform their decisions.
Google has designed these new experiences to highlight and drive attention to publishers.
An example of people preferring human content over AI content is reviews. Are you really going to book your next family holiday based on an AI recommendation that hasn’t travelled or experienced the city first-hand? Or will you take exercise advice from a robot that doesn’t lift weights?
For most of us, the answer is no — or, at least, not our preference.
People want to read content, especially purchase decision content written and tested by authors with experience. And Google is well aware. It’s why it rolled out its Helpful Content Update.
It doesn’t make sense that Google shouts from the rooftops for content from experts with first-hand experiences and then lets AI take over.
Reason #3 – Google Ads Revenue
According to Oberlo, Google makes 58% of its annual revenue, or $162 billion, from Google Search ads.
A massive reason why Google makes so much money is because businesses are paying to put their websites at the top of Google Search.
Companies won’t pay that much to be placed below the fold under the AI snapshot.
Not only that, but another 11% of Google’s revenue ($32.7 billion) comes from Google Display Network Ads. These are the ads that people bid on through Google to be placed on high-traffic websites.
People will stop bidding on these ads if these websites stop receiving traffic.
Reason #4 – Some Queries Won’t Be Covered By AI
Google has said they’ll be some queries SGE won’t answer.
This can be searches with information gaps where SGE doesn’t feel confident in a response. But also queries where Google knows that users want a human touch.
SGE: What’s Next, and Final Thoughts
Google wants to remain the number one search engine and to do that, they’ll roll out SGE methodically so they don’t lose user’s confidence.
According to the New York Times, Google plans to make all SGE features available to one million users in the US in May 2023. That number will increase to 30 million users by the end of 2023.
Google is also very aware of the limitations of generative language models, such as misinterpretations, hallucinations and bias.
Limitations aside, it’s important now more than ever to double down on SEO and content. In Google’s search guidance documentation, they noted that they reward high-quality content however it’s produced.
So Google will take away some of your traffic but is also giving you free reigns to use AI to scale your content production. Some SEOs have been incredibly aggressive in using AI in SEO.
While I believe AI has a place in helping creators, I’m more interested in getting in those top three positions in the snapshot. This is where first-hand experience produced by humans with expertise will be crucial.
While marketers spin AI content at scale, we’ll slowly produce high-quality pages that will be the source for AI while still leveraging AI in the creative process.
In my experience of using generative AI tools for the last 18 months, they haven’t necessarily saved time in my workflow. If anything, it has increased my standards to focus more on research, media and interviews to create content with proprietary value.
This is a moment for all of us to level up and create helpful content that AI can’t replicate. If you can do that consistently, you’ll mitigate traffic loss.
