How can I appear in Google's AI Overviews?
AI Overviews are not yet available on the search engine in France. However, you can anticipate their arrival by optimizing the content of your website. Indeed, appearing in Google's AI Overviews ensures a certain visibility and could prove to be a real lever for generating more traffic.
AI Overviews: what are we talking about?
AI Overviews, formerly SGE for Search Generative Experience, mark thearrival of generative AI in the search engine. These are text inserts that appear when a user makes a query. What's new? A detailed answer combining several search results, formulated in natural language.
Google uses its Gemini model, drawing information from different websites to reduce the time it takes a user to find the information they're looking for.
Recently, a clear reduction in the occurrence of AI Overviews has been observed. While these inserts were absent from 10% of search results in February 2024, they were absent from over 70% in June of the same year! This could mean that Google is more cautious than when the feature was launched.
Nevertheless, on August 15, 2024 Google announced the availability of this feature in six new countries : the UK, India, Japan, Indonesia, Mexico and Brazil.
At present, AI Overviews still rarely appear in results pages, but are becoming increasingly widespread around the world. So it's useful to understand the mechanisms that trigger their appearance, so you can optimize your content accordingly.
8 best practices for optimizing your articles for AI Overviews
Google seems to be giving mixed signals when it comes to best practice for inclusion in AI Overviews. Indeed, prior to the roll-out phase, the content used to generate responses was not necessarily part of the top 10 SERP results. We could therefore imagine that the conditions for being quoted in these inserts differed from classic SEO rules.
The trend seems to have changed, as it now turns out that the top SERP results are also the most frequently used content in the detailed responses generated.
Applying the rules of classic SEO
- Comply with E-E-A-T criteria
Starting in 2019, Google will use several criteria to assess the quality of content published online. In December 2022, the announcement of a new experience-based criterion completes the E-E-A-T concept.
E-E-A-T stands for "Expertise", "Experience", "Authoritativeness" and "Trustworthiness". In concrete terms, Google will use four indicators to feed its algorithm for ranking search results in the SERP.
Here are a few questions to ask yourself to determine whether content meets E-E-A-T requirements:
Expertise: Does this article bring something to my reader? Are the sources used reliable? Is it sufficiently rich, clear and understandable?
Experience : Does the author of my publication have enough concrete experience to tackle this subject? How can I promote it?
Authority: Is the author of this content entitled to speak on this subject?
Trustworthiness: Do readers have all the information they need to carry out a transaction? Are their personal data protected?
It's important to understand that Google relies on a global vision of the web, enabling it to identify links between authors, opinions, textual or multimedia content. This interpretation is made possible by metadata.
- Entering metadata
What is metadata? According to theInstitut de l'Information Scientifique et Technique, "Metadata can [...] be defined as a structured set of data used to define or describe a resource, regardless of its medium (paper or electronic)".
In other words, they enable content to be described in a language that search engines, in particular, can understand.
How is metadata useful for SEO?
Google relies on metadata to interpret a web page. The addition of this information is therefore recommended, as it not only enables the search engine to directly understand whether it is relevant to highlight a page in the SERP, but also to use the information provided in rich snippets.
It's possible to test a page's compatibility with the display of enriched results on desktop and mobile devices.
- Positioning on long-tail queries
The long-tail refers to a set of words, usually forming a proposition, and focusing on a specific search intent. For example: "Why fill in metadata?"
Although this search is rarer than "metadata" or "add metadata", long-tail keywords are less competitive and capture not only more traffic, but also targeted traffic.
Here are some examples of how to turn short-tail keywords into long-tail keywords:

- Add relevant links between your pages
Creating a network of logical links between the different pages of a website is essential. This is known as internal linking. This technique facilitates navigation for Internet users and improves the site's natural referencing by guiding search engines to the most relevant content.
👉 Read our article on the subject to learn step-by-step how to create a good internal link.
- Producing quality content
Google states that it favors content designed for users, not for improving a website's SEO scores. So it's vital to ask yourself what information your target audience is looking for, and what answers you can provide.
The more informative, structured and enlightening your articles, the more likely you are to be rewarded by the search engines.
Criteria specific to AI Overviews
Semji has carried out an in-depth study of how AI Overviews are triggered. In a webinar, its experts reveal several essential tips for increasing your chances of being quoted in AI Overviews.
- Add summaries to article headers
Web pages that effectively provide the answer to a question posed by an Internet user are given greater prominence. Conversely, if an answer is buried within a long piece of content, it's less likely to be cited. This is because LLMs don't need to explore an entire page, and find the answer more easily.
It can therefore be advantageous to add summaries at the top of your content so that informational queries trigger the display of your article!
- Rewrite articles in language that is clear and understandable to LLMs
A similar principle applies to the language used in articles. It's important to deliver useful information using simple words in clear sentences. It's easier for Google's LLMs to use this type of content, rather than extracting information and having to make it intelligible by generating a new sentence.
Long, convoluted sentences and the recurrent use of metaphors, for example, can be obstacles.
Tip: once the optimizations have been made, you can have Google explore the page via Search Console, so that any enrichments are quickly taken into account.
- Keeping an eye on AI Overviews
Finally, we recommend keeping a close eye on developments in AI Overviews. Their triggers have already changed between February and June 2024, and Google should be bringing new updates soon.
At present, the queries most likely to bring up these inserts are mostly related to the healthcare sector on long-tail keywords. What's more, they often begin with "How to", "Best" or "Can", enabling Google to directly generate answers to these questions.
The future of SEO in the age of AI Overviews
Finally, this latest update to Google's search engine follows in the footsteps of its predecessors, and once again aims to renew the user experience. However, the arrival of AI Overviews is a further step towards transforming search engines into answer engines. Users must continue to be cautious and critical of the results generated.
Google continues to announce new features for AI Overviews, which could indicate that worldwide deployment of these inserts is well and truly planned for the coming months. Thus, anticipating the potential arrival of AI Overviews in France could give Google a head start over its competitors.
Are you interested in this topic?