AI Search Optimization Tips From SEO Experts in 2026

According to the Semrush Sensor, AI Overviews appeared above organic search results for more than 48% of the keywords Semrush tracks in August 2026.

That’s a huge number, especially when you consider that Semrush’s database contains more than 28.8 billion keywords. Of course, this is Semrush’s keyword database, not the entire internet. But the trend is hard to ignore.

I keep seeing site owners talk about organic traffic drops that never really recovered. My website is no exception.

2025 was the best year for my website so far. The SEO content strategy I put together generated over $45K in income. Back then, my website was getting around 8–10K organic sessions a month, according to Google Search Console.

Those days are gone.

In 2026, I moved my old domain to Behind Rankings. Around the same time, I saw a pretty painful drop in organic traffic, partly because AI Overviews started taking away clicks that would have previously gone to websites like mine.

These days, my website’s organic traffic fluctuates between 1–3K sessions a month, according to Google Search Console.

While my organic traffic keeps dropping, my visibility in AI search is actually growing. Here’s what Semrush’s AI Visibility Toolkit shows for my website over the past few months:

I’m obviously happy to see my overall AI visibility going up, especially when I see historically strong pages like my AI SEO tools and AI marketing tools posts getting cited.

But there’s a problem: AI visibility can help create awareness, but that doesn’t always translate into clicks.

I recently had a conversation with Kanishka, a content manager at beehiiv, my favorite newsletter platform, about the current state of SEO and what brands can actually do to improve their visibility in both traditional and AI search.

That conversation got me thinking that I wanted to hear from more people who actually work in SEO, SaaS, B2B, and more. So I reached out to other experts and asked what’s been working for them.

The strategies below come directly from people working in the industry and are based on their own experience.

Some are classic SEO tactics you’ve probably heard before. Others are more specific to how AI systems understand brands, content, and what people say about you across the web.

And no, none of these strategies can guarantee that your brand will show up in AI Overviews or LLMs. I don’t think anyone has AI search completely figured out yet.

But if I were trying to improve a brand’s visibility in AI-generated search results right now, these are the strategies I’d be paying attention to.

My experience boosting AI visibility

I’ll start with a few observations from my own website and some experiments I’ve run over the past few months.

One thing I noticed pretty quickly is that Google seems to love fresh content when it comes to AI Overviews. My latest LinkedIn posts are often pulled into AI Overviews, and Behind Rankings’ LinkedIn account is also frequently cited for searches about my brand.

About two weeks after launching the Behind Rankings Instagram account and reaching 1,500+ profile views, Google seemed to connect the account to my brand. Soon after, I started seeing my Instagram posts appear in AI Overviews for branded searches.

Publishing an About Behind Rankings page had a similar effect. It started showing up in AI Overviews for brand-related queries almost immediately.

Gemini cited almost the entire About Behind Rankings page when generating a response about my brand, which supports my earlier assumption that having a dedicated page about your brand with the most frequently asked questions may help your website, rather than third-party sources, get cited in LLMs and AI search results.

There are also a few things that still work surprisingly well for traditional search.

Updating obsolete content. When I add information around emerging search queries, even keywords with very little search volume, I can see impressions and clicks increase.

I still believe in building pillar pages that properly cover a topic and satisfy search intent. But one of my recent experiments made me rethink how I approach emerging topics.

Instead of adding a new topic as another section of a big pillar page, I started publishing shorter posts targeting emerging keywords. Some of them ranked surprisingly quickly and were also picked up by AI Overviews.

For example, this newsletter on the topic of Semrush MCP and Claude performs pretty well in search results even though this was never meant to be a properly optimized content piece.

Another interesting pattern: some of my posts that ranked in Google’s first search result page last year, and some that still do, are now appearing in AI Overviews even when they’ve dropped out of the top 10.

My AI SEO tools, AI marketing tools, and Substack alternatives posts are good examples.

Not everything worked, though.

After the domain migration, I experimented with fixing entity signals and adding FAQ schema. Honestly, I didn’t see any noticeable improvement from either.

And several months after the migration, I still see LLMs cite my old domain. 😕 This makes me think AI systems may need considerably more time than Google to update what they know about a website.

I’ve also seen some interesting results from third-party local businesses’ websites.

Strong internal linking and well-connected content clusters have helped improve impressions, clicks, and average position. I’ve also seen AI-generated content perform well on third-party sites, but it wasn’t just AI-generated content sitting there on its own. Those pages also had internal links, expert quotes, and clear business information.

So, from my experience, AI visibility is about implementing an omni-channel approach. It seems to come from consistently building a stronger presence across your website, social profiles, and other websites, while continuing to do the SEO basics that still work.

How a Botswana safari company boosted AI visibility

Nonofо Joel, the CEO at Brandesis, shared a case study with me from Wild for Game Safaris, a photographic safari company based in Botswana.

The company offers private wildlife safaris across destinations including the Okavango Delta, Moremi Game Reserve, Savuti, Nxai Pan, and the Central Kalahari. Its trips focus on hands-on safari experiences, with local guides, private tents, meals prepared in the field, and wildlife photography.

When Joel started working with the company in February 2025, the website was getting only around 10 clicks per month from organic search results on Google. It also had no AI citations and had not yet gone through any meaningful SEO work.

Over the next 18 months, the team worked on the site’s technical SEO and built connected content clusters around the company’s first-hand knowledge and expertise.

The clusters focused on topics such as:

Botswana wildlife
Photo safaris
Mobile safaris
Safari destinations
Safari planning

According to Joel, by August 2026, the website was ranking for around 1,426 organic keywords and generating an estimated 907 monthly organic visits.

But the AI visibility results caught my attention even more. In August 2026, Semrush’s AI Visibility tool showed that 81 pages on the website were cited in AI search results. Those pages generated 161 citations and 2 brand mentions.

I think this case study makes an important point for smaller websites: you don’t need to be a huge publishing brand to get visibility in AI search. Instead, you need to consistently build a topical authority in your niche and showcase first-hand experience when creating content.

How Triumph grew AI assistant sessions 7x

Triumph’s Global SEO & AI Search Manager, Filippo Danesi, and fellow Semrush Ambassador shared with me what happened after the company started treating AI search as a separate acquisition channel.

Triumph is a global lingerie brand founded in Germany in 1886 and now headquartered in Switzerland. The company operates in more than 120 countries, selling underwear, sleepwear, and swimwear.

According to Filippo, sessions from AI assistants that can be tracked in Google Analytics grew about 7x year over year after Triumph introduced a dedicated AI search strategy and its own KPI.

So, what did they change?

Every month, the team ran the same set of prompts across different AI models and tracked two metrics separately:

How often Triumph was mentioned
How often Triumph’s own pages were cited

Triumph appeared in roughly one in three AI answers, but its own pages were cited much less often.

That gap was important. It showed the team that AI answers about Triumph were often based on third-party content rather than the company’s own website.

According to Filippo, these things made the biggest difference:

1. They organized their content plan around prompts

Instead of relying solely on traditional SEO tools for keyword research, the team looked at the queries AI models use when answering shopping-related questions.

In their niche, many of these queries followed patterns like “[topic] reviews reddit.” They used these queries to decide what topics to cover and how to structure their headings.

According to Fillipo, their editorial pages now receive AI citations at roughly 3.4 times the rate of their commercial pages.

2. They started reporting AI search traffic separately

The team started tracking AI search traffic separately in GA4.

They used a simple regex to identify referrals from AI search tools and tracked this data each month alongside their organic search traffic.

This made the 7x growth in AI search traffic easier to spot. Once the results were clear, it also became easier to get internal support for investing more in editorial content.

As Filippo put it:

“Results nobody can see don’t get budget.”

I really like this case study because it separates two things marketers often combine when talking about AI visibility: brand mentions and page citations.

In my experience, mentions show whether AI talks about your brand. Citations show whether AI uses your content as a source. In other words:

𝗠𝗲𝗻𝘁𝗶𝗼𝗻𝘀 = brand visibility
𝗖𝗶𝘁𝗮𝘁𝗶𝗼𝗻𝘀 = content visibility

A mention means your brand, website, product, or person is referenced in the AI-generated answer. A citation means the AI system uses your page as a source and often shows a link or source card to your website.

I believe these are two different metrics, and they should be measured differently if you want to get a clear and detailed overview of your brand performance in AI search results and LLMs.

So, keep an eye on this when choosing a suitable AI search visibility tracking tool.

How Pretty Little Fit improved AI search visibility

Phoo, an SEO strategist at Transcendence Creatives Asia, shared this case study from Pretty Little Fit, a Malaysian activewear brand that focused on AI search optimization strategies.

The goal was simple: make the brand more visible across Google AI Overviews, Gemini, ChatGPT, and visual search.

Therefore, the team focused on the following areas:

1. Answer-first content

According to Phoo, the biggest change was how articles were written. Rather than opening with a long introduction, every page started with a direct answer to the user’s question, such as:

“The best socks for Pilates are non-slip grip socks with silicone grips on the sole.”

Only after answering the question did the content expand into benefits, comparisons, and product recommendations. This format makes it much easier for LLMs to extract and cite the answer.

2. FAQ schema

Each page included structured FAQ markup to reinforce key questions and answers.

Instead of hiding important information deep in the page, the content surfaced it in a format that both search engines and AI models could interpret more easily.

3. Multimodal visual SEO

The team also treated images as a ranking asset. They optimized:

descriptive file names
clear alt text
product schema
high-quality product photography
images designed to rank in Google Images and visual search

This mattered because customers were discovering products through Google Images, Pinterest, drag-and-drop image search, and AI tools like Gemini and ChatGPT.

Within three months, Pretty Little Fit went from having virtually no search visibility to appearing on page one of Google and earning prominent placements in Google AI Overviews.

Monthly visits increased by 146%, rising from 102 to 251, alongside growth in both citations and brand mentions.

One of the examples came from the keyword “grip socks for pilates in Malaysia.” The brand’s Pilates grip socks achieved two valuable outcomes at the same time:

ranked on the first page of Google Images
were cited inside Google AI Overviews for the same product-related query

That combination helped generate product visibility and customer inquiries from both Google Images and AI search results.

This case study shows that for ecommerce brands, product pages need to clearly communicate:

what the product is
who it’s for
what makes it different
where it can be used
how it compares with alternatives

Pretty Little Fit is one of the clearest examples in my research where AI visibility and organic traffic increased together.

That doesn’t happen in every case. Some brands gain AI mentions while organic clicks stay flat, and others see the opposite.

Tips from the SE Ranking team

Bogdan Krupin from SE Ranking shared one of the frameworks I found most useful in this discussion.

He looks at AI brand visibility as three connected layers:

How you position your brand
Where your brand stands in competitive searches
What people and communities say about you

And all three need to tell a similar story.

For example, your website might position your product as the best solution for a specific audience. But if reviews, Reddit threads, partners, or other websites describe your product differently, AI systems may pick up those conflicting signals and give a less consistent answer.

As Bogdan pointed out, your official brand positioning shouldn’t fight against what people are saying about you. These signals should support each other and help AI systems build a clear picture of your brand. This becomes especially important when people search for your brand directly.

An AI system might use information from your website, but it can also look at reviews, comparison pages, community discussions, partner websites, and “best of” lists. So, you can’t fully control your AI visibility from your own website.

You also need to pay attention to what the rest of the web says about your brand and make sure that information is accurate, relevant, and consistent.

AI search optimization tips for SaaS

Omid Ghiam, Head of Growth at Gumloop, shared a few AI search optimization tips specifically for SaaS companies.

The first tip from Omid was about authorship. Instead of publishing everything under vague bylines like “Company Team,” SaaS companies should put real experts behind their content.

I completely agree with this.

A software review written by someone who has never actually used the product is not nearly as useful as a review from someone who tested it and has an opinion based on that experience. The same goes for technical guides, product comparisons, and advice on how to implement a tool.

A real author can explain what they tested, what surprised them, what worked, and what didn’t. That kind of first-hand experience gives readers something a generic AI summary simply can’t.

Omid also talked about net new information gain, and I think this is one of the most important AI search optimization tips in this entire discussion.

AI systems are already very good at finding and summarizing information that exists online.

So if your article repeats the same points as 20 other articles, it might be perfectly accurate, but there’s not much that makes it stand out. This is where first-hand experience becomes really valuable.

You can add things like:

Original research
Product testing
Customer examples
Screenshots
Personal experiments
Expert interviews
Specific results
Your own opinions
Detailed implementation notes

But simply adding a sentence like “In my experience…” to an otherwise generic article isn’t enough. The experience should be visible throughout the piece.

That’s why I included my own observations earlier in this article. I’m not just explaining what AI visibility is. I’m sharing what I actually saw after tracking it across my website and social profiles.

Omid also believes fully human content is stronger than content that is mostly AI-generated, especially in competitive niches. I’d add one important nuance here.

Using AI for research, outlining, or editing doesn’t automatically make content bad. The problem starts when the final piece has no original thinking, experience, or clear point of view.

AI can help you produce content faster. It can’t replace the person who actually knows what happened, what changed, and what the results meant.

AI and LLM citation tips for agencies

Scott Stockdale from Avium shared a useful case study from his agency work.

One of his clients saw a 98% increase in clicks since March 2026. Another moved from around position 12 to position 2 for a keyword the team had been working on since January.

And the interesting part is that there was no secret AI SEO trick behind these results.

The team focused on things that are actually pretty familiar to SEOs:

Better on-page targeting
More relevant H1s, title tags, and URLs
FAQs added to service pages
Internal links from service pages to relevant blog posts
Consolidating articles targeting the same keywords
301 redirects
Moving video content higher up on blog pages

I think this is a good reminder that traditional SEO still matters for AI visibility. AI tools still need to discover, understand, and evaluate your content.

If your page has a confusing title, weak structure, several articles covering the same topic, and no internal links, you make it harder for both search engines and LLMs to understand what your website is about.

Scott’s example also shows why internal linking is so important (and I totally agree with this point).

When a service page links to relevant blog content, you create a clear connection between the commercial page and the informational content.

That helps search engines understand how the pages relate to each other and which topics your website has expertise in.

The same idea applies to AI search. Your website should make those relationships easy to understand. For example, a content cluster around AI search optimization could include:

AI search visibility
Google AI Overviews
Generative engine optimization
Brand mentions in AI search
AI visibility tracking
Expert-led content
Original research

You don’t need to force links between every article. Just link to related content when it actually helps the reader and makes the connection between topics clearer.

Scott also talked about consolidating articles that target the same keywords. This is especially relevant for agencies managing larger websites where content can quickly pile up.

If you have several pages covering essentially the same topic, you can end up splitting signals between them. It can also become unclear which page should rank in Google or be cited by an AI system.

Sometimes the best AI SEO move isn’t creating another article, but figuring out which existing page should own the topic and consolidating or redirecting the rest.

Can Reddit improve AI search visibility?

Nikita Vlasyuk, CTO and co-founder of FeedHeat, shared an interesting observation about Reddit.

According to Nikita, building a presence on Reddit helped FeedHeat get cited by AI models in relevant subreddit discussions. That’s exactly why I started contributing to existing subreddits from my personal account and started a subreddit for Behind Rankings. In my experience, you have to be constantly active and contribute regularly. Otherwise, it could take more time for Google and LLMs to associate your Reddit activities with your brand.

Reddit is full of the kind of information AI systems can use when answering recommendation and product-related questions:

First-hand experiences
Product comparisons
Honest opinions
Specific use cases
Questions from people trying to solve real problems

Your website usually tells people how you want your brand to be seen. Reddit can show how real people actually talk about your brand and your product.

That community perspective can become another signal AI systems use when deciding how to describe a company or recommend a product.

Reddit is not the place to copy and paste your marketing messages.

If you want to build a presence there, start by listening. Read the discussions. See what people are asking. Pay attention to the problems they’re trying to solve.

Then contribute when you actually have something useful to say. You don’t need to mention your brand in every answer.

The goal is to become a helpful part of the conversations your audience is already having.

Where you should start

After everything I’ve tested and learned from others, I don’t think AI SEO is that different from traditional SEO.

The basics still matter, including creating helpful content, clear page structure, internal links, search intent, and mentions of your brand across relevant third-party websites.

However, AI search adds another layer on top of your efforts to boost organic rankings. And staying on top of emerging AI SEO trends helps you adjust as things shift

What you say about your brand should match what customers, partners, and communities say about you. And your website content should offer something more than a rewritten version of what’s already out there.

Even though I’ve been implementing most of the tips from this list, my own results have been somewhat mixed.

Organic traffic went down while AI visibility went up. Updating old content helped, but some shorter articles performed much better than I expected. On the other hand, fixing entity signals and adding FAQ schema didn’t make much of a difference. And months after the domain migration, I’m still seeing LLMs cite my old domain.

So, I’m definitely skeptical when someone promises a guaranteed way to get into AI Overviews or LLMs.

I don’t think anyone has AI search completely figured out yet.

For now, I’d focus on the things you can actually control, such as creating helpful content, sharing your real experience, building a credible brand, and making sure the information about your brand is consistent across the web.

FAQ

What is AI search optimization?

AI search optimization is the process of improving how often and how accurately your brand appears in AI-generated search results.

This can include Google AI Overviews, ChatGPT, Perplexity, Bing AI experiences, and other generative search tools.

AI visibility can mean different things. Your brand might be:

Mentioned in an AI answer
Cited as a source
Recommended for a specific use case
Associated with a particular topic or entity
Shown for branded or non-branded queries

One important thing to remember: AI visibility doesn’t always lead to clicks.

Someone can see your brand in an AI answer, remember it, and never visit your website. So AI visibility can still build awareness even when it doesn’t generate direct traffic.

How can I improve AI search visibility?

I suggest starting with the basics.

Create expert-led content, share original insights and first-hand experience, make your brand positioning clear, build strong internal links, and get mentioned on relevant third-party websites.

It also helps to monitor where and how AI systems mention your brand so you can see what’s actually working.

Does traditional SEO still matter for AI search?

Yes. Traditional SEO still helps AI systems discover, understand, and evaluate your content.

But AI search adds another layer. Your visibility can also depend on things like who is associated with your content, what people say about your brand, community discussions, and information published on third-party websites.

Can Reddit improve AI search visibility?

It can.

Reddit discussions can give AI systems additional context about your brand, especially when people share real experiences, opinions, comparisons, and use cases.

But Reddit isn’t a place to drop promotional messages everywhere.

If you want to build visibility there, focus on being genuinely helpful and contributing to conversations that are relevant to your audience.

Do I need to publish more content?

Not necessarily.

Publishing more isn’t always the answer. In many cases, updating existing content or creating a few focused supporting articles can work better than publishing lots of generic content.

I’d focus on covering gaps, answering emerging questions, and adding information that isn’t already available in every other article.