All of our content is researched and written by human writers. No robots here!
If you want to learn how to optimise content for AI search, the first thing you need to know is this: Your audience ain’t sitting around waiting for your blog to explain the basics anymore.
More and more of them are asking ChatGPT, Google’s AI Overviews, Gemini, Copilot, and other AI search engines to do that first.
And the numbers are now big enough that this isn’t some niche “early adopter” thing you can put on the “ignore, not a big deal” list. It’s already shaping what content shows up in AI search results and what doesn’t.
Don’t believe me? Let’s slap down some cold, hard facts from the McKinsey people:
- Half of consumers now intentionally use AI powered search
- Around 50% of Google searches already include AI summaries
- Unprepared brands could see traffic declines of 20% to 50% from traditional search channels (gasp!)
So no, content hasn’t suddenly become irrelevant. But its job has changed quite a bit.
Your website content now needs to support visibility, influence, and usefulness all at once: AI visibility, influence before the click, and usefulness after it.
I’m gonna’ tell you how to do that in this blog. But first, let’s get clear on all the ways AI has changed people’s search habits:
People are researching like crazy before they get to your site
People are using more conversational, natural language when they’re searching for stuff online. Less “best CRM software” and more “best CRM software for a small recruitment agency with a tiny team and no patience for complicated setups.”
Which, honestly, makes sense. People are asking AI tools to compare, summarise, sense-check, shortlist, and explain things quickly before they commit to clicking around ten tabs on traditional search engines like they used to.
And that shift is showing up in the data:
- Just 8% of one- or two-word searches generated an AI summary (Pew)
- 60% of searches beginning with question words triggered an AI summary (Pew)
- 80% of search users rely on AI summaries for at least 40% of their searches (Bain)
What that means in real life is this: by the time someone lands on your site, they probably already know the basics. They’ve even ruled out three competitors and formed opinions about your category without visiting a single homepage.
Which means your content needs to meet people later in the journey than it used to.
Why your content might be getting fewer clicks
If you’ve seen your organic traffic dipping, it doesn’t automatically mean your content is bad. It could be because AI summaries are answering more of the low-stakes informational stuff.
That “what is X,” “how does Y work,” and “what’s the difference between A and B” traffic is easier to satisfy without a click now.
On some search engine results pages, people are getting the direct answers they want before they ever visit a site, especially when they’re still in the skim-and-sense-check phase.
Again, the numbers back this up:
- When an AI summary appeared, users clicked a traditional search result just 8% of the time (Pew)
- About 60% of searches now end without the user progressing to another destination (Bain)
- Organic web traffic is being reduced by 15% to 25% as consumers lean more on AI-written results (Bain). McKinsey puts the risk even higher for unprepared brands, at 20% to 50% (McKinsey)
So yes, some clicks are disappearing. But in many cases, that’s a behaviour shift, not proof that your content strategy sucks all of a sudden.
The traffic you do get may be more qualified than before
Now for the less depressing section of this blog. (Woohoo, finally!)
Fewer clicks doesn’t always equal worse traffic. No siree bob! AI seems to be filtering out some of the casual browsers and sending through people who are more informed, more intentional, and more likely to convert.
I’m not saying every AI visit means people buy immediately with tears in their eyes, I’m just suggesting that the people who click through are probably arriving with more context and clearer intent.
So if your team is still judging content performance mostly by raw pageviews, this is your nudge to stop doing that. Or at least to stop doing only that.
Now for the good stuff. How to make sure your site shows up in those coveted AI spots.
AI is handling the first round of research. Your site needs to handle the next one.
AI tools are increasingly taking care of the “give me the basics” stage, so your content has to work harder in the next stage: helping someone decide whether you’re a good fit.
That means more content that helps people compare, evaluate, trust, and move. Content that goes that extra step further, offers more substance, and helps people make decisions.
So what should brands be creating more of now?
- Detailed service pages that actually explain what you do, who it’s for, what’s included, and what happens next.
- Comparison pages that help people weigh up options, approaches, packages, or categories.
- “Who this is for/not for” content that helps people self-qualify instead of forcing them to guess.
- Case studies and proof-driven pages that show genuine outcomes, not just big claims in a nice font.
- Pricing, process, and FAQ content that answers the awkward questions people absolutely have but probably won’t ask on a sales call.
- Expert-led thought leadership with actual opinions, original examples, and first-hand experience.
Because here’s the thing: generic top-of-funnel explainers are very easy for AI to summarise. Specific, experience-backed, brand-distinct content is much harder to replace.
AI can’t easily replicate your perspective, your process, your proof, your weirdly specific edge, or the examples you have from doing the work in real life.
That’s where your content gets more much more valuable for people.
Give your content a better chance of showing up in AI answers
internet has been reborn. Answer engine optimization. Generative engine optimization. AI search optimization.
Strip all that back, and the job is pretty much this: make your content easy for AI systems to access, understand, trust, and cite.
The best way to do that is to create unique, satisfying, people-first content. Google said that themselves. This isn’t really a separate universe from good search optimization, more of an evolution of it.
So, practically speaking:
Make it accessible
Don’t hide important information in PDFs, image-heavy graphics, tabs no one can crawl properly, or clever design choices that look nice and say absolutely nothing to a machine. Keep key content indexable and in the text on the page. Google explicitly recommends making sure important content is available in textual form and crawlable.
Make it easy to understand
Use descriptive headings. Answer obvious questions directly. Build pages in a way that directly answers user queries and makes sense to both humans and AI models.
Make it easy to quote
Comparison tables, lists, bullet points, concise summaries, FAQs, and well-structured service content all help because they make your information easier for AI to scrape and reuse accurately.
Make it trustworthy
Show who’s behind the content. Keep pages updated. Use original examples. Support claims with evidence. Link between related pages. Use structured data and schema markup where it actually helps clarify what’s on the page. If your site looks anonymous, vague, outdated, or half-finished, don’t expect answer engines or search engines to treat it like a wise sage.
Make it better than the summary
AI can summarise. Your page still needs to reward the click. Give the reader more depth, more nuance, more proof, more examples, and a clearer next step than the summary could ever provide on its own. That’s super important whether someone found you through AI powered search engines or through more traditional search results.
In other words: don’t write content that can be fully replaced by a quick skim in an AI box.
Put more time into off-page SEO
Because AI systems pull from such a massive mix of sources, improving your own website pages is only part of the job. Here’s why:
- Your own site may account for just 5% to 10% of the sources AI search uses (McKinsey)
- AI search pulls from affiliate links and user-generated content (McKinsey)
- Wikipedia, YouTube, and Reddit were among the most commonly linked sources in both AI summaries and standard search results (Pew)
That means you can have a stunning, well-written, technically solid website and still barely show up in AI answers if other sites on the wonderfully wide web don’t really mention you, review you, compare you, quote you, or discuss you.
So if your brand only really exists on your own channels, that’s now likely going to impact search visibility.
Your website is obviously still a very very big deal in all of this, don’t get me wrong. But it’s just not the only thing you need to be concerned about now if you want more visibility. That might mean:
- Improving your presence on review platforms.
- Earning mentions in credible third-party publications.
- Getting your experts to publish smart, quotable thoughts on LinkedIn.
- Encouraging customer-generated content where it makes sense.
- Making sure your business facts are consistent across listings, profiles, and directories.
- Giving people something worth referencing in the first place.
If the wider web says very little about you, or worse, says random inconsistent things, you’re leaving a lot of your visibility up to chance.
You need a new way to measure content performance now
If the journey has changed, the scorecard has to change too. I know, I know. This blog has been pretty stat-heavy. But I think the way search behaviour is changing is kinda’ fascinating (nerd, and proud). So here’s more:
- Just 16% of brands systematically track AI search performance (McKinsey)
- Traffic from AI features is still reported in the main Performance report in Search Console (Google)
- Microsoft launched AI Performance in Bing Webmaster Tools, including metrics like citation counts, cited pages, grounding queries, and visibility trends over time (Microsoft)
So alongside rankings and traffic, you should also be looking at:
- Qualified traffic
- Engaged sessions
- Assisted conversions
- Branded search lift
- Leads from decision-stage content
- Referral traffic from AI tools
- Citation frequency in AI generated responses
- Off-site share of voice and brand presence
You’ll also want to keep a closer eye on referral sources in Google Analytics, especially if you’re trying to understand where AI traffic is actually landing and what it does next.
Because a drop in top-of-funnel clicks might sit right beside an increase in better-fit visitors, stronger branded demand, and more visibility inside the places that now influence decisions.
This is not the end of content marketing. It’s a change in what good content has to do.
AI search has changed how people find content, size it up, and decide whether it’s worth their time. But it hasn’t changed the core truth underneath all of this: high-quality content still wins.
Maybe even more so now. The difference is that “good” has a higher bar. It needs to be original. Specific. Well explained. Properly structured. Full of actual value.
The kind of content that helps someone move forward, not just skim the basics and leave. That’s not a new idea for us. It’s the whole reason The Content Lab exists.
Want help creating content that stands up in search, shows up in AI answers, and gives people a reason to stick around once they land? Get in touch with head honcho Abby to see how we can help.
Facebook
Twitter
LinkedIn
