SEO is Dying. AI Engine Optimization is Here. 6 Facts.
Google's expansion of AI mode into Chrome will accelerate the death of the blue links. Here's why AEO is such a critical part of your growth strategy.
Is SEO dead?
We know AI assistants are growing in popularity - 700/800 million weekly active ChatGPT users. This is about 10% of the world’s adult population.
We know clicks are disappearing. A study from Seer Interactive analyzed 10,000 keywords and found that organic CTR plummeted by 70% when AI overviews were present, falling from 2.94% to 0.84%. It’s estimated that between 60% and 70% of searches now result in no clicks. Oh my.
A quick pause. The biggest error I see with hot takes about SEO vs. AEO (AI Engine Optimisation) is about search volume vs. clicks. Let’s be clear. AI will increase search volume. People will ask more questions, which is great for Google. Their products like AI Max are monetising intent vs. clicks. What’s disappearing are clicks.
Google’s latest product launch is going to expand that trend. Their expansion of AI mode into Chrome will mean a lot more search is done within AI vs. users searching via the blue links. It also means if they click on a web page, instead of reading the web page, they can just ask Gemini questions about that web page. They don’t even have to read it.
AEO is an incredible new channel to generate visibility for your brand, product and services.
Here’s 6 key facts about AEO.
1. Authority Still Rules
SEO is dying in its current form, but it’s not dead :)
OpenAI indicated, as part of their GPT-5 model launch, they’d increasingly look to external sources for data and focus their efforts on reasoning models to reason against that data. We’ve seen referral traffic from AI assistants like ChatGPT increase sevenfold from 2024 to 2025. ChatGPT, on average, appears to use external sources in over 54% of its queries.
It means AI engines still rely on Bing/Google results to fetch source links before synthesizing/reasoning.
It means transactional SEO, e.g., appearing for buying-related keywords, is still a worthwhile investment as those links get pulled into AI assistants’ answers.
Good quality SEO -> Google quality AEO.
2. Citations Matter
Much like SEO, citations still matter, from community and trusted sources. I covered how each LLM favors different sources in this article . A lot of it’s based on who those LLM companies are partnered with.
Reddit is become a dominant force in AEO. It achieved 40.1% of all large language model citations and is the most cited source across AI-generated search results.
Companies like Ramp are hiring professional Redditors to gain more visibility in community sites that have an out weighted influence on the LLMs.
ChatGPT heavily favours authoritative sources, with Wikipedia and major news outlets accounting for 54% of all its citations. PR is making a return.
Google AI Overviews cite blog content 46% of the time, meaning that blogs no longer get traffic but are used for surfacing answers in AI overviews.
In summary, citations about your product and service really matter.
3. Enable the Bots or Lose the Game
Much like traditional search engines, AI engines use specialized bots to discover and index content, but with a fundamentally different purpose.
Instead of crawling to build ranked search results, AI crawlers collect data to train models and retrieve real-time information for citations.
OpenAI operates three distinct crawlers: GPTBot for training future models, OAI-SearchBot for real-time ChatGPT search results, and ChatGPT-User for on-demand content access during conversations.
AI bots now account for up to 20% of total web crawling activity and behave differently than Googlebot. They prioritize textual clarity over design elements and focus heavily on structured content with clear headings.
Unlike traditional SEO where you optimize for ranking positions, AEO requires optimizing for two scenarios: getting your content into training datasets for long-term knowledge, and structuring pages for real-time retrieval when users ask specific questions.
AI crawlers specifically hunt for citation-worthy content that can directly answer user queries. To do that, you need to make sure these bots can crawl your website.
4. Structure Your Pages
As we said above, AI bots heavily favour sites with structured pages for real-time user answers. Unlike human visitors who scan visually, AI crawlers read your content linearly and rely on clear semantic signals to understand what information to extract and cite.
The most effective structure follows a simple hierarchy: use H1 tags for main topics, H2 for key questions or sections, and H3 for supporting details. AI engines particularly favour FAQ formats, Q&A sections, and content that directly answers conversational queries like “How do I...” or “What is the best way to...”. Research shows that pages with clear heading structures are 3x more likely to be cited by AI engines than pages with poor information architecture.
5. Write for Questions, Not Readers
AEO is different from SEO. Asking an AI assistant isn’t the same as searching Google for keywords.
AI models prefer Q&A structured content because it maps directly to user prompts.
Instead of single product pages optimised for 3 to 5 keywords, you’ll need many niche versions of your product page that perfectly answer all variations of questions users can have about that product/service.
We’ve consistently seen improvements in AI visibility at HubSpot as we’ve rolled out a strategy that takes advantage of his tactic.
6. You Can *Kind* of Measure It
There’s been an explosion of rank trackers for AEO. They all have access to the same data and all measure the exact same things.
Here’s the challenge with current rank trackers:
- They can’t replicate your memory. When searching on ChatGPT, the results will be personalised for you. These rank trackers can’t replicate your results.
- There are differences between asking ChatGPT a question on your desktop and asking it the same question via the API. The AEO tools are bulk asking questions via the API.
- It’s impossible to know how customers are querying ChatGPT and other AI engines about your product or services. It’s so different from searching via keywords. How do you predict how customers speak about you to other people.
There is still merit in using these tools. You’ll get a ‘best guess’ at how your visibility is improving, so you can at least correlate the impact of your work to a ‘share of voice’ metric going up within the LLMs.
Once ChatGPT adds a paid advertising product to target freemium users, we’ll start to get much better data.
No marketing strategy in 2026 is complete without some focus on AEO. Google’s roll-out of AI model will accelerate it’s importance. The fun part is, it’s so new, the playbooks are still being figured out.
Until Next Time,
Happy AI’fying
Kieran