The problem
Something is quietly changing about how people search.
A growing number of consumers are typing questions into AI tools, ChatGPT, Gemini, Perplexity, and others, and receiving a direct answer rather than a list of links. They are asking things like:
“What is the best project management software for a small team?”
“Which accounting platform is easiest to use for a freelancer?”
“Can you recommend a sustainable skincare brand in the UK?”
In response, the AI tool produces an answer. It may mention a handful of brands. It may cite a few sources. It may recommend a product. And then the customer decides based on what they were told.
If your brand is not mentioned, you are not in that conversation.
At the same time, Google has introduced AI Overviews, which now appear at the top of many search results pages. These summaries are generated by AI and pull from sources across the web. Again, the brands and pages that are cited are visible. The ones that are not cited are not.
Businesses need a repeatable way to see whether they appear in AI answers, how often, in what context and how that compares with competitors.
Why this matters now
This is not about AI replacing Google overnight. That is an oversimplification. Google remains the dominant search platform, and traditional SEO continues to matter. But the way people interact with search is changing alongside it.
For products that involve research and comparison, an AI answer can become one step in the customer journey. Businesses can measure whether they appear in those answers alongside their existing search metrics.
The consequence is that visibility must be understood differently. A brand could rank on the first page of Google for important terms and still be entirely absent from the AI-generated answers that a proportion of its potential customers are relying on. These are two separate visibility questions, and right now most businesses are only measuring one of them.
What businesses misunderstand
Before talking about solutions, it is worth clearing up some common misunderstandings.
“If we rank on Google, we will automatically appear in AI answers.” Not necessarily. Ranking well is a useful signal, but AI tools do not simply mirror Google rankings. They draw on multiple sources, interpret content differently, and make their own judgements about which information is relevant and trustworthy for a given question.
“Adding schema will guarantee that AI tools cite us.” Schema and structured data help machines understand what your pages are about, which is genuinely valuable. But no markup guarantees a citation. It improves the clarity of your signals; it does not control the output.
“One prompt test is enough.” AI answers can vary between sessions, between tools, and over time. Testing your brand name once in ChatGPT on a Tuesday morning gives you a snapshot, not a reliable picture. Systematic testing across multiple queries, tools, and time periods is needed.
“If ChatGPT does not mention us once, we are invisible everywhere.” Different AI tools behave differently. Perplexity may cite your brand where Gemini does not. Google’s AI Overviews may use your content where a conversational AI tool does not. Visibility needs to be measured across the landscape, not judged from a single test.
“This is only a problem for big brands.” Small and mid-sized businesses can also benefit from measuring AI visibility. Review whether your content clearly explains what you offer, answers relevant customer questions and is accessible to the search systems you want to reach. Small and mid-sized businesses may be more exposed. Larger brands with strong authority signals are often better represented in AI training data and source selections. Smaller brands need to work harder to ensure their content is clear, accessible, and trustworthy.
What the evidence says
Research into what has become known as Generative Engine Optimization, or GEO, is still at an early stage. Early academic research published at the ACM KDD 2024 conference explored how websites can be measured and potentially improved inside AI-generated answers. The researchers found that certain content characteristics, including citing authoritative sources, writing clearly, and providing specific and useful information, appeared to be associated with higher visibility in AI-generated responses.
Google Search Central is clear that helpful content should be created for people first, that it should be accurate, and that it should demonstrate expertise and trustworthiness. These principles, which sit at the heart of Google’s evaluation of quality, also appear relevant to how AI systems select and surface content.
OpenAI’s crawler documentation explains how its web crawlers work and how website owners can allow or restrict access using their robots.txt file. This means that if a website is blocking AI crawlers, either intentionally or by accident, those pages may not be accessible to AI training and retrieval systems at all.
Google’s structured data documentation explains how markup can help Google understand page content and support eligible rich results. For e-commerce brands, Google Merchant Center guidance makes clear that product quality, including accurate titles, prices, and availability, directly affects how well products are understood and surfaced.
Taken together, the evidence suggests a practical approach: make content clear, accurate, accessible and useful, then measure whether visibility changes. This gives search and AI systems better information to work with, while citations remain uncertain.
What good AI visibility looks like
Good AI visibility is not a single score or a guaranteed position. It is a combination of signals and a process of ongoing measurement.
In practical terms, it may include:
Your brand being mentioned in relevant AI answers when a customer asks a question in your category
Your product pages or content being cited as a source in AI-generated responses
Your website being accessible to AI crawlers, not blocked by accident
Your schema and structured data being complete and accurate, so machines understand what your pages are about
Your product information being clear, up to date, and well-structured in your feeds
Your content addresses the questions customers are asking
Your visibility being tracked consistently over time, not just tested once
Your competitors being monitored so you can see where they appear and you do not
The goal is not to game AI tools. It is to be a clearer, more trustworthy, more accessible source of information, for both people and machines.
What does this mean in practice
Here is a practical example.
Imagine a marketing manager at a mid-sized UK software company. Their product helps small businesses manage their finances. A potential customer sits down and types the following into an AI tool:
“What is the best accounting software for a small business in the UK?”
The AI tool produces answers. It mentions two or three products by name. It may explain what each one is good for. It may link to review pages or the products’ own websites. The customer reads the answer and forms an opinion.
Now the marketing manager needs to ask:
Does our brand appear in that answer?
Which competitors are mentioned?
What sources is the AI drawing on?
If our brand is missing, why might that be?
Is our website clear about what we do and who we serve?
Is our content answering the questions customers are asking?
Are our product pages structured in a way that machines can interpret correctly?
Is our site accessible to AI crawlers in the first place?
These are not abstract technical questions. They are practical business questions that directly affect whether a potential customer hears about your company before they decide.
Without a way to measure these things systematically, the marketing manager is flying blind.
How NeuraCite helps
NeuraCite is an AI visibility and citation intelligence platform. It helps brands understand how they appear inside AI-generated answers and what can be done to improve those signals.
It helps answer questions such as:
Are we mentioned when customers ask AI tools about our category?
Are our competitors being mentioned instead of us?
Which sources are AI tools drawing on for our industry?
Is our website accessible to AI crawlers?
Is our Schema complete and accurate?
Are our product feeds ready for AI-powered shopping and discovery?
What content gaps are making it harder for AI systems to understand what we offer?
Is our AI visibility improving over time?
NeuraCite helps brands measure where they stand, identify what can be improved, and track whether changes are having an effect. It covers prompt testing across AI tools, schema and structured data audits, product feed readiness, crawlability checks, competitor visibility analysis, and content recommendations.
NeuraCite helps improve the signals that AI systems may use. It does not control AI outputs, guarantee citations, or promise specific results. No platform can do that honestly. What it can do is give your team the visibility and insight to make better decisions.
Practical next steps
If you are new to this area, here are some straightforward steps to start with.
Test your own brand. Open ChatGPT, Gemini, and Perplexity. Type the questions your customers are most likely to ask. See whether your brand appears.
Test your category. Ask AI tools for recommendations in your sector, without mentioning your brand. See which companies are mentioned and which are not.
Check your competitors. Note which brands appear consistently across multiple tools and queries. Understanding where you stand relative to competitors is as important as understanding your own position.
Review your website content. Is it clear what you do, who you serve, and what makes you different? Is it written to be genuinely useful, or mainly for search engines?
Check your schema. Are your pages using structured data correctly? Is it up to date and complete?
Check your crawl access. Is your robots.txt file accidentally blocking AI crawlers? This is a simple technical check that can have a meaningful impact.
Audit your product feeds. If you are an e-commerce brand, are your product titles, descriptions, prices, and availability accurate and complete?
Track over time. A single test tells you very little. Set up a process to monitor your visibility regularly, so you can see whether things are improving.
Limitations and uncertainty
It would be wrong to end this article without being honest about what is still unclear.
AI-generated answers are not fully predictable. Different tools behave differently. The same tool can give different answers to the same question on different days. Model updates, changes in training data and shifts in how queries are interpreted all affect what gets surfaced.
No platform, including NeuraCite, can guarantee that your brand will be cited by any AI tool. Schema improvements, better content, and cleaner product data can all help AI systems understand your brand more clearly. But there is a difference between improving your signals and controlling the outcome. The former is achievable. The latter is not.
GEO, as a formal discipline, is still emerging. The academic research is at an early stage and continues to develop. Practices vary across the industry, so findings should be treated as a developing picture rather than settled rules.
The honest position is this: businesses that improve the clarity, quality, and accessibility of their content and data are giving AI systems more to work with. That is a reasonable and measurable goal. The results should be tracked carefully over time, with realistic expectations.
Sources used
Google Search Central, Creating helpful, reliable, people-first content: guidance on the principles of quality content that search and AI systems are designed to reward
Google Search Central, How structured data works: documentation on structured data markup and how it helps machines understand page content
Google Merchant Center, Provide high-quality data: guidance on product data quality and why accurate, complete product information matters
OpenAI, Overview of OpenAI crawlers: documentation explaining GPTBot, OAI-SearchBot, and how website owners can manage AI crawler access via robots.txt
Aggarwal et al. (2024), “GEO: Generative Engine Optimization”, KDD 2024: the foundational academic paper introducing GEO as a framework for measuring and improving brand visibility in AI-generated answers