Redefining SEO: AI Overviews and the road ahead
The SEO landscape is changing fast, and many of the tactics we rely on today are becoming obsolete. To keep driving organic traffic and real value, we need to rethink our approach.
New developments – like Google’s AI Overviews, rolled out in May – are reshaping search results, shifting us away from the old “10 blue links” model.
AI is influencing not just platforms and processes but also how users behave and interact with content.
This article covers key insights and practical strategies to help you adapt to these changes and prepare your SEO efforts for 2025 and beyond.
1. Having a featured snippet may translate into an AI Overview citation
When you search for “what is a zap in Zapier,” you may notice that the featured snippet appears as a citation in the AI Overview.
This is not an isolated case.
Many reports highlight similarities between featured snippets and the answers provided in AI Overviews.
Some experts have even suggested that this redundancy could lead Google to eventually phase out featured snippets altogether.
What this means for SEOs
There’s still value in trying to appear for featured snippets. Think of this as one way to have a higher chance of appearing in the AI Overviews.
If you appear in a featured snippet you are increasing your opportunity to appear in AI Overviews.
Now is the time to get your ducks in a row and build a process to optimize for featured snippets. This will continue to be useful, should Google get rid of the featured snippets in the future.
Dig deeper: AI SEO: How to be visible in Google AI Overviews, chatbots, LLMs
2. Each citation in AI Overviews can have multiple sources (URLs)
Here’s an example from searching for the same query “What is a zap in zapier.” A click on the link icon shows two URLs that were used in creating this part of the answer.
About 21.1% of queries trigger overviews with each answer containing an average of 8.9 links, per a study by Rich Sanger and Authoritas.
This means we have 8.9 links in the AI Overviews plus the organic results in SERPs.
What this means for SEOs
AI Overviews aren’t here to destroy organic traffic. Instead, they are reshaping the ranking landscape and creating new opportunities in SERPs by providing more visibility.
Now is the time to shift our focus: rather than solely aiming for a spot in the top 10 results, we should also prioritize appearing in AI Overview citations.
The good news is that you don’t need to rank in the top 10 to be cited in these overviews.
3. Differences between SERP results and pages selected for LLMs
There have been various studies showing the differences in the top 10 ranking results and the URLs cited in AI Overviews.
Up to 46.5% of the URLs included in AI Overviews rank outside the top 50 organic results, per an Advanced Web Ranking study.
What this means for SEOs
Focus on creating high-quality content, regardless of whether you’re in the top 10 rankings.
Even if you don’t rank highly, you still have the potential to be cited in AI Overviews or by large language models (LLMs).
Low-quality content is becoming obsolete, and there’s increasing value in producing meaningful content beyond just SERP rankings.
While measuring this value may be difficult at the moment, we will soon have the tools to assess it more effectively.
Dig deeper: The art of AI-enhanced content: 8 ways to keep human creativity front and center
4. Brand mentions are important
We hear about backlinks losing their value every year. LLMs and AI Overviews are shifting that narrative once again.
The more brand mentions you have (even without a link), the greater the likelihood that your content will appear in answers within AI Overviews and other LLMs.
What this means for SEO
Backlinks remain important, but their significance is growing even more.
Getting your brand mentioned is crucial for increasing your chances of appearing in LLMs and AI Overviews.
Dig deeper: How to harness the power of brand mentions across the search universe
5. Be present in all the sources Google uses for information
Reddit comprised about 4% of the training data for GPT-3, yet it held a significant weight of 22% in its influence.
Similarly, Wikipedia made up around 1% of the training data but has three times that weight in its impact on LLMs.
Given its current importance, it’s clear that being active on platforms like Reddit is essential for maximizing your Google visibility.
What this means for SEOs
To appear in AI Overviews and other LLMs, it’s crucial not only to be mentioned or cited but also to actively establish your brand on platforms like Reddit.
Instead of simply driving traffic from Reddit to your website, consider posting relevant information directly on the platform.
When done thoughtfully and without spamming subreddits, this approach can resemble a white or gray hat strategy of parasite SEO that adds value to your business.
Additionally, maintaining a presence on Wikipedia is important – just remember to avoid spamming.
As Rand Fishkin put it, “You have to be present with native content in the zero-click platforms because that is the future.”
Dig deeper: How to win with generative engine optimization while keeping SEO top-tier
6. Learn from CTR tests
We can no longer afford to lose clicks because our titles in SERPs are not optimized for CTR.
If some of the clicks are going to be taken by AI Overviews, then we need to make every impression in SERPs count. There’s no better way to do this than optimizing your CTR.
What this means for SEOs
I can’t emphasize this enough: SEO is evolving, and our strategies must adapt accordingly.
CTR tests should be a fundamental aspect of our approach, as even small adjustments in click-through rates can lead to significant results.
These improvements can be observed at various levels, including individual pages, folders and sitewide performance.
7. Guard your brand against manipulation
Negative SEO, which once involved building toxic backlinks to competitors’ websites using spammy anchor texts, is set to evolve.
Competitors may now attempt to manipulate AI Overviews and LLMs to present negative information about a brand.
What this means for SEOs
To protect your brand from manipulation, monitor brand mentions and build a strong identity that is not easily undermined.
We need to develop effective ways to measure our brand’s health in SERPs and online.
While we currently have metrics such as the total branded searches, clicks, share of voice and interest in the brand (as indicated by Google Trends), we should also consider creating additional metrics and methods to assess our brand health comprehensively.
Dig deeper: Online reputation management: Top 10 hurdles and how to overcome them
8. Learn about RAG
As Michael King puts it, “Search, as we know it, has been irrevocably changed by generative AI.”
Answer bots like Perplexity and AI Overviews use retrieval-augmented generation (RAG), a technology that leverages live data to ground answers in facts and reduce hallucinations.
What this means for SEOs
Just as we once learned how Google operated, we must now understand the workings of RAG technology.
By investing time in grasping how RAG functions, you can better navigate this new reality and adjust your tactics and strategies accordingly.
Many SEOs have yet to change their playbooks simply because they don’t fully understand these changes. Take the time to learn about RAG – it’s crucial for adapting to the future of SEO.