Search is changing. Instead of ten blue links, we now get AI-powered summaries, conversational responses, and sometimes… no need to click at all. For content creators, SEOs, and product owners, the game hasn’t ended – but it’s definitely shifted.
Google and ChatGPT are both surfacing content in new ways. The rules are still evolving, but if you want your content to show up where users are now looking, there are concrete things you can do.
Here’s what we know so far – and how to work with it.
When Impressions Rise but Clicks Don’t
One of the more interesting shifts in AI-powered search is what’s now being called “The Great Decoupling”. Google has confirmed that as AI Overviews roll out, many sites are seeing more impressions but fewer clicks.
At a 2025 Search Central Live session, Google’s Martin Splitt explained that impressions might spike due to the expanded visibility from AI Overviews—your content may be referenced or displayed more often—but users are less likely to click, especially when their question is already answered in the summary. That said, Google notes that the clicks you do get tend to be more qualified, often leading to deeper engagement or conversions.
This changes how SEO teams need to interpret performance:
- A rise in impressions without a matching rise in traffic isn’t a failure—it may just reflect better surface-level visibility.
- Instead of optimizing for CTR alone, focus on engagement, conversions, and post-click behavior.
- Monitor your Search Console closely for sudden impression spikes, especially from long-tail or question-based queries that might be triggered by AI Overviews.
Common Misconceptions About AI SEO
Before you dive into tactics, it’s worth clarifying what AI SEO is not about. A few myths persist that can lead teams in the wrong direction:
- There’s no separate AI ranking algorithm. Google doesn’t treat AI Overviews as a new ranking system—it uses the same quality signals and page evaluation processes as it does for traditional search.
- You don’t need to write “for AI.” If your content is readable, well-structured, and actually answers a question, it already fits what AI systems are looking to pull.
- AI results don’t replace standard SEO. Google still sends massive traffic through standard blue links. Optimizing for AI Overviews is additive—not a replacement.
Clearing up these points helps focus efforts on what actually moves the needle.
How to Optimize for Google AI Overviews
Google’s AI Overviews (formerly SGE) are essentially auto-generated summaries that pull from various web pages to answer user queries directly. If your content is solid, it can be quoted or linked within those summaries. But there are rules.
Start with the basics:
- Make sure your site is crawlable (don’t block Googlebot accidentally)
- Avoid noindex, nosnippet, or overly strict meta tags if you want to appear
- Serve your pages cleanly (200 status, no weird redirects)
Google’s documentation keeps repeating the same phrase: “helpful, reliable, people-first content.” That means original content that genuinely addresses user intent – not reworded listicles. AI Overviews favor depth and uniqueness.
Also worth noting: as of mid-2025, AI Overview and AI Mode impressions and clicks are counted in Google Search Console under “Web Search” – but they’re not separated out. You won’t see an “AI traffic” filter (yet). You’ll need to infer from query patterns, especially long-form and conversational ones.
And don’t forget the page experience side. Google has said bad UX (slow loads, cluttered design) can keep you out of Overviews, even if your content is relevant.
So: fast pages, original content, crawlable structure, and no snippet-blocking tags. Nothing groundbreaking – but more important than ever.
How to Optimize for ChatGPT Results
ChatGPT is now pulling from the web in real-time. When users ask questions, it can respond with live citations and links – yours included, if you’ve set things up right.
The main thing? Don’t block OpenAI’s crawler. Specifically:
- Allow OAI-SearchBot in your robots.txt (this bot indexes pages for ChatGPT’s responses)
- Don’t block Bingbot either – ChatGPT’s browsing uses Bing under the hood
It also helps to understand what these bots do:
- OAI-SearchBot = shows your page as a citation in live ChatGPT responses
- GPTBot = used to train OpenAI models (your content goes into the model, not the search index)
So if you want traffic, let OAI-SearchBot in. You can block GPTBot if you don’t want your content in training data – OpenAI supports this split setup.
As for the content itself: ChatGPT favors clear answers, well-structured formatting, and current information. FAQs, guides, and product pages with clean structure tend to perform best. If you’re in e-commerce, OpenAI is also testing product recommendations – merchants can opt in by feeding structured data or allowing indexing.
And when ChatGPT does link to you? It tags the URL with utm_source=chatgpt.com, so you can filter and track that traffic easily in GA or other tools.
Tracking and Measuring AI Search Traffic
This part’s still messy. Google lumps AI traffic into regular Web Search. ChatGPT gives you referral tags but no volume indicators. Still, here’s how to keep an eye on things:
- Google Search Console: AI traffic is included, but not labeled. Watch for spikes in long-form queries or changes after SGE rollouts.
- SurferSEO AI Tracker: One of the few tools built specifically to monitor AI visibility. It tracks:
- When your content appears in AI Overviews
- AI-specific CTR trends
- Positioning comparisons vs traditional results
- When your content appears in AI Overviews
- Analytics filters: Look for utm_source=chatgpt.com in your reports. That’s your signal a user came from a ChatGPT citation.
- Server logs: Want to know when AI crawlers are indexing your site? Watch for hits from OAI-SearchBot, ChatGPT-User, or Bingbot.
Also useful: track what those users do once they land. Session length, bounce rates, conversion – early data suggests AI-driven visits might be smaller in volume but higher in quality.
Optimizing for Google’s AI Search Results (SGE & AI Mode)
AI Overviews are just one layer. Google’s AI Mode adds conversational follow-ups and deeper answer threads. It’s a more dynamic interface – but the playbook stays mostly the same.
A few principles worth sticking to:
1. Content still matters more than anything.
Unique, detailed content wins. Pages that address questions in full, show expertise, and go deeper than surface-level answers are more likely to be pulled in.
2. Technical SEO can still block you.
Slow sites, bad markup, broken schema – all still hurt your chances. Structured data helps Google understand your content better, but only if it’s accurate.
3. Snippet visibility is a lever.
Want to be quoted? Don’t use nosnippet. Want to stay out of AI summaries? Use max-snippet or limit crawling.
4. Format with machines in mind.
Headers, semantic HTML, and clean structure make it easier for AI to extract relevant chunks. Treat AI like a very fast reader looking for clarity.
5. Watch the query landscape evolve.
As users adapt to AI search, they’ll ask more complex, long-tail questions. Adapt your content to follow these trends – answering niche, high-intent queries could pay off more than competing for generic ones.
🎯 What’s Next: Shopping in ChatGPT
ChatGPT is leveling up its commerce game. According to OpenAI’s help doc, when users indicate buying intent (e.g., “best hiking boots under $100”), ChatGPT now:
- Displays product carousels featuring title, images, and prices pulled from third-party structured metadata, not paid ads.
- Generates simplified descriptions and highlights user-friendly labels like “Budget-friendly” or “Most popular,” based on customer reviews and aggregated ratings.
- Includes review summaries and star ratings sourced from third-party providers—though not verified by OpenAI, they enrich the shopping experience.
- Leverages intent cues: If the user specifies budget or features (e.g., color, size), ChatGPT uses those preferences to tailor the carousel.
- Allows deeper merchant integration: OpenAI is inviting merchants to submit product feeds directly—this will make listings more accurate and timely.
What this means for you: If you’re in e-commerce, optimizing structured product data (e.g., schema, clean metadata, updated pricing & images) and allowing OAI‑SearchBot to crawl your pages increases your chances of being featured in these AI shopping carousels.
Final Thoughts
There’s no secret trick to winning in AI search. But there is a mindset shift: you’re writing for humans, and for machines that summarize content for humans.
Focus on quality, structure, and technical accessibility. Let the right bots in, measure what you can, and adapt as the tools and user behavior evolve.
The goal is the same as it’s always been: earn trust, answer real questions, and make the content useful. The only difference now is – sometimes, it’s an AI doing the reading.
