How to optimise articles for AI Overviews

AI Overviews (AIO) are all what people are talking about - SEO is dead again. SEO has been dead since my first BrightonSEO in spring 2017.

AIOs are the AI-generated summaries that show at the top of the Google SERPs. Gone are the days of featured snippets (although I saw one or two!) - AIOs are all that we SEOs are talking about. This info is pulled from across the web - you don’t even need to rank on the first page to be featured here. If your content isn’t structured in the right way, Google might skip over it entirely, even if it ranks well traditionally. However, important caveat - keep an eye on your SERP rankings as you optimise for AIO. AIOs fluctuate, traditional SERPS are still the backbone of search engines.

In 2025, as part of my role as an SEO Specialist at GoDaddy, I optimised over 50 articles with AIO in mind. A sample analysis of the results showed 100% of reviewed articles ranking #1 in AI Overviews, with 254 keywords reaching Top 3 positions. Here's what I did — and what I learned along the way.

A quick but important caveat

Before diving in, it's worth being honest about something: these results are backed by GoDaddy's strong domain authority, long-established online presence, and global brand recognition. Applying the exact same methodology to a small local business would likely be a much slower, harder process, and results would vary significantly. I know this first-hand, having previously worked agency-side with mostly local Romanian small businesses.

That doesn't make the approach any less valid. It just means that if you're working with a newer or smaller site, you should set realistic expectations and think of AIO optimisation as a long-term investment rather than a quick win.

Start with a benchmark - look at what was already ranking in Google

Before changing anything, I did a thorough SERP analysis - both for us and for our competitors. The question I was trying to answer was simple: what do the articles that appear in AI Overviews have in common?

I used SEO tools, but I also did a lot of visual analysis, looking at the search results page with my own eyes. I was interested in all the YouTube videos that ranked in the AIO, too, as we were also growing our channel.

Having a benchmark is important, as you can refer to it 6 months-1 year down the line and see if you need to reverse anything or keep going!

How to improve articles to reach AIO

1. Headings became questions

One of the first things I noticed was that articles ranking in AIO tended to use question-based headings. Instead of "FQDN Explained", the heading would be "What is a FQDN?". It sounds like a small change, but it makes a big difference. Google's AI is trying to answer questions, so content that's already structured as questions and answers is much easier for it to use.

I went through articles and rewrote headings to mirror how a real person would actually search for something, using Ahrefs and SEMrush data where possible.

2. FAQs at the end

You might have noticed that FAQ sections don't really appear as a special feature in Google search results anymore. But that doesn't mean they're useless. I found from my research that FAQs are one of the main ways content gets pulled into AI Overviews.

I added relevant FAQ sections at the end of articles, including comparison questions (like "URL vs FQDN - what's the difference?", “What's the difference between a virtual address, physical address, and P.O. box?”) which tend to be exactly the kind of thing people search for when they're trying to understand a topic properly.

A real-world example that really stuck with me: I noticed that for a particular topic, an Australian article was consistently outperforming its American counterpart in US search results. Both articles were fresh, hreflang was implemented correctly across both, and there were no obvious technical issues (no rogue noindex tags, no redirects, nothing glaring). The content was essentially the same, adjusted only for local pricing, grammar, spelling conventions, and a few localised examples (like a .com example vs a .com.au one).

The difference? The Australian article had eight FAQs at the end. The US article had none.

So I added some of the Australian FAQs and some US-specific ones, and the gap closed. It's a small thing that had a disproportionately large impact and it was interesting to see how “just” the FAQ section was such a differentiator.

3. Tables and bullet points for structured information

Just like Featured Snippets, AI Overviews love structured content.

If information can be presented as a table or a bullet point list rather than a long paragraph, that makes it much easier for Google to extract and use. It’s also easier for your readers to skim through the article (cause no one is reading anything in full anymore). So I reformatted sections of articles - comparisons, step-by-step processes, lists of features.

4. Clear, no-nonsense definitions

This one was a significant shift in approach that needed a guideline sent to all our writers. A lot of the articles I was looking at improving had a very personal, conversational style - think sentences like "So you're wondering what a domain is" or anecdotes about client experiences. Charming, but not what Google's AI is looking for.

AIO tends to favour content that gets straight to the point, like clear definitions, direct answers, no unnecessary fluff. That doesn't mean the content has to be boring, but it does mean front-loading the important information rather than building up to it.

5. Fresh publish dates and lots of internal linking

Google favours fresh content, so where articles had been updated, reflecting that with a new publish date helped. I also added a lot of internal linking, connecting articles to each other so that Google could understand the relationship between topics and see the site as a reliable, interconnected resource rather than a collection of isolated pages.

The results

After applying these changes across 50+ articles, the results were striking. In a sample analysis, 100% of reviewed articles were ranking #1 in AI Overviews, with 254 keywords reaching Top 3 positions in traditional search results too. So optimising for AIO didn't come at the expense of regular rankings. The two go hand in hand more than you might expect.

What you should do

If you manage a website or work in marketing and you're trying to get your head around AI Overviews, here's the simple version:

Google's AI wants to answer questions clearly and quickly. If your content is structured to do the same (with question-based headings, clear definitions, FAQs, and well-organised information) you're giving it exactly what it needs.

You don't need to be an SEO expert to make these changes. A lot of it is just about thinking like your reader: what are they actually trying to find out, and have you answered it as clearly as possible?

Good luck optimising for AIOs

The SEO world moves fast, and AI Overviews are still evolving. But the core principle behind all of this isn't new: write content that genuinely helps people, structure it so it's easy to read, and make sure it answers the questions people are actually asking. Google's AI is, in many ways, just getting better at rewarding exactly that.

And if you're a small business owner reading this wondering whether any of it applies to you - it does, just with patience. The fundamentals are the same. The timeline is just a little longer.