aio
Don’t Fear The Bots – Why AIO Can’t Rank Without You
Published: 23/09/25 - Updated: 23/09/25
The AI Landscape
Two sides of AI Search – Google (Gemini/AI mode/AI Overviews) and LLM search (ChatGPT, Perplexity, Claude).
ChatGPT’s growth has been exponential and is expected to be 1% of all web traffic by 2025. This might seem like a small amount, but when you think of the sheer volume of all site traffic, it equates to [insert figure] daily sessions.
This might be an optimistic view of its growth, but it shows the momentum behind AI search and its potential.
This is furthered by the news around OpenAI and Perplexity looking to create their own browser, potentially eating into more of Google’s market share.
AIOs
Currently we’re seeing an increase in zero click searches (with 60% of all searches in the UK and US ending without a click). This alongside increased ads, more enhanced SERP features, AI overviews and AI mode mean this is only going to get worse and traditional listings are getting pushed further and further down the page.
Since September last year there has been a 500% increase in the number of AIOs and it’s estimated that these now reach 1.5 billion users each month, and this is only going to continue to grow as they begin rolling out across more and more search queries.
- 17% of all searches in the UK and US now display AI Overviews
- 99.5% of AIOs pull data from at least 1 top 10 search result
- 35% drop in position 1 CTR when AIOs are present
Why AIOs and AI Mode are essential for organic success
Search is fundamentally changing. With the changes in AI search, Google is no longer the only port of call for search. Alongside this, user behaviour is changing and conversational queries are becoming more normalised as users move away from keywords to prompts and longer form searches. Ultimately, this means performance is no longer just about clicks, with brand playing an increasingly important role.
For PR teams looking to close the gap between ambition and reality, the key steps are clear.
- Start by assessing where AI and data could add strategic value rather than just tactical efficiency.
- Integrate these tools into planning to guide campaigns with evidence rather than intuition alone.
Aligning PR strategy with insight-driven decision making allows businesses to improve visibility, authority, and influence while managing risk in an environment where public perception can shift quickly.
Prompts Vs Keywords
| Element | Keywords | Prompts |
| Length | 1-5 words | 10-25 words+ |
| Intent | Search focused | Conversation focused |
| Context | No context | Wider context |
| Example | ‘Used family cars’ | I’m looking for an automatic second hand car that has enough space for the family, but isn’t too large or expensive. What makes and models are available? |
Ultimately, AI Overviews still need to pull information from somewhere, so we want you to be that source of information and source of truth.
LLM Search
Mindshare and brand visibility are more important than ever when it comes to AI visibility, and it’s important to be found as we know it converts well. Typically, while LLM traffic makes up <1% of site traffic, users are highly engaged and it converts really well, with session durations 5x longer and conversion rates 12x higher than organic search.
Leveraging AI optimisation
In order to leverage AI search, you first need to understand how it works and how LLMs are finding their data.
All LLMs use a technique called fan-out search in order to build up an idea of what content is available for a topic and what is most relevant. This takes the form of several smaller sub-searches. For example, if you’re looking for information on “How can I naturally lower cholesterol without medication?”, LLMs might search for content around:
- Natural alternatives to statins
- Alternative treatments for high cholesterol
- Plant sterols and stanols cholesterol
- Cholesterol-lowering supplements evidence
- Lifestyle changes to improve cholesterol
- Cholesterol-lowering foods NHS
This gives it vital context to best answer your query, and essentially does the typical user journey you might take from traditional search, but does it in seconds and summarises all of the available data.
Now LLMs won’t provide this data by default, but you can ask for it – simply ask ChatGPT or Perplexity what types of fan-out searches were made for your prompt and it will give you a breakdown of what it searched for and why. Using this data can be fantastic for reverse engineering what type of data the LLMs are looking for, and the type of content you might need to create in order to appear. Combine this with the sources that are cited in the result and you can build up a picture of what good content looks like.
You can then use this data to also map brand entities – understanding how relevant your site is for various topics and sub-topics helping you map out where you have content gaps or good content depth.
This is all only as useful once you can track visibility – using various tools or simple dashboards, tracking the number of brand mentions or citations for various prompts you want to appear for is crucial to measure progress and check that you are on the right track.
When it comes to Gemini search (AI Mode and AI Overviews) Google recently provided a breakdown of all the key elements that might help you appear in AI Search. This includes making sure you:
- Focus on unique, valuable content for people – follow Google’s helpful content guidelines
- Provide a great page experience – ensure good UX, useability, and site speed
- Ensure Google can access your content – ensure crawlability and technical requirements (e.g. robots.txt and sitemaps)
- Manage visibility with preview controls – use nosnippet, data-nosnippet, max-snippet or no-index tags to set preferences
- Make sure structured data matches the visible content
- Go beyond text for multimodal success – use multiple content formats such as images or video
- Understand the full value of your visits – evaluate site engagement and iterate
- Evolve with your users – adapt content to the way users are searching (targeting longer tail queries/prompts)
Conclusion
Ultimately AI is giving us the freedom to think more creatively. It can do the grunt work faster and more efficiently allowing us to scale and grow at a much greater pace. But there will always be the need for human guidance. AI is great at generating base content at scale, but needs a human to validate the emotional resonance. AI is great at coding and generating schema markup, but lacks perfect accuracy and so needs a human to verify. It can suggest relevant related pages and internal links, but these might not align with a good user journey.
When optimising for humans, you will always need the human touch so don’t fear the bots!
Stay visible in the age of AI search. Kaizen helps brands optimise for AI Overviews, LLMs, and evolving search journeys to drive measurable growth. Get in touch today.
Ed Coles
Ed is the SEO Lead at Kaizen, with 8 years experience in the SEO industry. Having worked across various agencies from small to large multinational independents, Ed initially started in the world of PPC before quickly moving to the world of SEO. From helping clients in the third sector and retail, to working on larger global ecommerce sites, Ed has been responsible for supporting the roll out and delivery of global SEO strategies in EMEA and NORA across some iconic brands such as Vans, Timberland, The North Face, LEGO, and Adidas, among many others.