- What are branded and unbranded queries, and why does the distinction matter?
- Is it really worth monitoring my brand's AI visibility, or is it still too early?
- What really happens when someone searches for your brand on Google with AI Overviews?
- How does brand visibility work on ChatGPT: mentions vs. cited links?
- How do you actually track branded and unbranded queries on ChatGPT?
- Three concrete scenarios: what changes for an e-commerce, a professional firm, and a manufacturing company?
- What tools exist to automate brand tracking in AI?
- What are the risks of not monitoring my brand's AI visibility?
- Why it's a hot topic now: is the AI market in Italy really ready for this kind of analysis?
- What mistakes are most common when you start monitoring AI visibility?
- How do I know if my brand's AI visibility is improving?
- So, is it worth starting branded vs. unbranded AI monitoring for your company?
- How we see it at SHM
- Frequently Asked Questions
- Branded queries (with your brand name) and unbranded queries (generic, category-based) call for different AI visibility strategies and should be monitored separately.
- On Google, AI Overviews appear in 73% of branded searches and 78% of non-branded ones: your site risks ranking below the AI answer even when people search for your own name.
- ChatGPT mentions brands 3.2 times more often than it cites them with a link: being mentioned matters as much as being linked, and it's measured differently.
- Monitoring works with a systematic test sheet across ChatGPT, Gemini and Perplexity: 30 minutes a week is enough to get useful data.
On ChatGPT, brands that appear in the internal query the system writes for itself before searching the web end up in the final answer 68.9% of the time. Those that are only found later, during the search, make it 2.1% of the time: a 33x gap.
What are branded and unbranded queries, and why does the distinction matter?
A branded query contains your brand name (e.g., 'Nike running shoes for the marathon'). An unbranded query is a generic category search (e.g., 'best running shoes'). On ChatGPT and other AI assistants, the two types lead to very different answers, and your brand may appear, or disappear, differently in each.
Imagine two potential customers. The first asks ChatGPT: "Is brand X reliable for SMEs?" The second asks: "What's the best management software for SMEs?" Both could buy from you. But the first already knows you, and the second doesn't.
For the first, the AI has to confirm or deny your reputation. For the second, it has to decide whether to name you among the options. These are two completely different challenges, and they call for two different strategies.
Understanding this distinction is the starting point for any serious SEO and AI visibility work.
| Query type | Example | What you measure | Goal |
|---|---|---|---|
| Branded | “Does Brand X also ship internationally?” | Accuracy and tone of the AI answer about your brand | Reputation and accuracy of information |
| Unbranded | “Best sustainable packaging suppliers” | Whether and where you appear among the options cited | Acquiring new customers |
| Branded + category | “Brand X vs. competitor Y for e-commerce” | Competitive positioning in AI answers | Defending your positioning |
Is it really worth monitoring my brand's AI visibility, or is it still too early?
With almost 19 million Italians using AI assistants and over 20 million exposed to Google's AI Overviews, the channel is already relevant for most companies. Not monitoring it means not knowing what is being said about your brand in a setting your customers use every day.
In August 2026, AI assistants reached 18.9 million unique visitors in Italy, more than double the roughly 9 million of January 2025, according to Sensemakers/Netcomm data published by Adnkronos. Penetration among the Italian digital population rose from 20% to 40% in less than two years.
Italians exposed to Google and Bing AI Overviews are already over 20 million (source: Sensemakers based on Comscore data). This means that even people who don't use ChatGPT directly get AI-generated answers when they search on Google.
| Your situation | Is it worth monitoring? | why |
|---|---|---|
| You have a recognizable brand in your industry | Yes | AI may already be answering questions about your brand, even inaccurately |
| You operate in a category with comparison searches (e.g. software, B2B services, e-commerce) | Yes | Unbranded category queries have high commercial value |
| You're a startup or new brand with no awareness | Partly | Focus on unbranded category queries to get known |
| You run a purely local, offline business | No, for now | The AI channel is still marginal for hyperlocal searches without digital intent |
What really happens when someone searches for your brand on Google with AI Overviews?
According to an Ahrefs study of 232.8 million crawls, in September 2026 Google's AI Overviews appeared for 73.3% of branded keywords and 78.3% of non-branded ones. In over three quarters of cases, the AI answer appears above the brand's own website.
The most surprising finding comes from Ahrefs: in 76.2% of cases, the AI Overview for a branded search appears in position 1, above the brand's official website. The average position of the AI Overview for branded queries is 1.49.
This means that when a customer searches for your name on Google, the first thing they see is not your website, but an AI-generated summary. If that summary is inaccurate, incomplete or favors a competitor, the damage is real, and you may not even know it.
As we discussed in our article on AI Overviews and brands, this dynamic profoundly changes how you need to manage your online reputation.
How does brand visibility work on ChatGPT: mentions vs. cited links?
ChatGPT mentions brands 3.2 times more often than it cites them with a link. Being mentioned without a link is already a meaningful form of visibility, but the two metrics should be measured separately because they affect user behavior differently.
According to BrightEdge, 44% of ChatGPT prompts contain no brand mention at all. But commercial queries, those with words like 'where to buy' or 'deals', generate 4 to 8 times more brand mentions than purely informational queries.
Another useful figure: across 100 different answers to the same question, ChatGPT mentions 44 brands in total on average, but names only about 10 in any single answer. In niche categories, 21% of mentioned brands are 'dominant', meaning they appear in 80% or more of answers (source: Search Engine Land).
The practical goal is to become one of those dominant brands in your category. As explained in our article on ChatGPT and primary sources, the quality and structure of your content is the main factor.
How do you actually track branded and unbranded queries on ChatGPT?
The most accessible method is a manual test sheet: a list of branded and unbranded questions to test every week on ChatGPT, Gemini and Perplexity, recording whether the brand appears, in what tone and in which position.
There's still no tool that does it all automatically, but the manual process is simpler than it looks. What matters is being systematic: same questions, same platforms, same cadence.
As we explained in our article on SEO tracking on Google and ChatGPT, AI monitoring requires a different approach from the classic check of Google rankings: here there are no fixed rankings, only answers that vary.
- 1 · Marketing managerBuild your query listWrite 10-15 branded questions (with your brand name) and 10-15 unbranded ones (category-based). Include comparison and commercial questions (‘where to buy’, ‘best for…’).
- 2 · Marketing managerChoose your platformsTest on ChatGPT, Gemini and Perplexity. They're the three platforms with the most users in Italy. Test in incognito mode or in a fresh session to avoid personalization.
- 3 · Marketing manager or agencyRecord the resultsFor each answer, note: does the brand appear? In what position (first, second, third)? With what tone (positive, neutral, negative)? Is it cited with a link or just mentioned?
- 4 · Marketing managerRepeat every weekAI answers change over time. Weekly monitoring lets you spot changes and link them to the actions you took (new content, PR, site updates).
- 5 · Marketing manager or agencyAnalyze the gapsCompare branded and unbranded presence. If you show up well on branded queries but not on category ones, the problem is AI awareness. If it's the other way around, the problem is AI reputation.
Three concrete scenarios: what changes for an e-commerce, a professional firm, and a manufacturing company?
Monitoring priorities change depending on the type of business: e-commerce should cover unbranded commercial queries, professional firms branded reputation queries, and B2B manufacturers comparative category queries.
Illustrative examples: numbers are meant to show the reasoning, they are not real customer data.
| Company type | Branded queries to monitor | Unbranded queries to monitor | Main risk | Priority action |
|---|---|---|---|---|
| Fashion e-commerce | “Is Brand X reliable?” / “Does Brand X offer free returns?” | “Best online shops for sustainable clothing” | Not showing up in category lists | Category content optimized for AI |
| Consulting firm | “Is Studio X specialized in…?” / “Studio X reviews” | “Best consultants for SME internationalization” | Inaccurate information about your specialization | Structured, up-to-date service pages |
| B2B manufacturer | “Does Company X also make custom components?” | “Italian suppliers of precision metalwork” | Not being cited in industry comparisons | Presence on industry sources cited by AI |
What tools exist to automate brand tracking in AI?
Tools dedicated to monitoring AI visibility are still under development, but some platforms already offer useful features. Comparing the options helps you choose based on your budget and the complexity of your brand.
| Tool / Approach | What it monitors | Indicative cost | Suitable for |
|---|---|---|---|
| Manual test sheet (Google Sheets) | Custom queries on ChatGPT/Gemini/Perplexity | Free (just your time) | SMEs with a limited budget |
| Semrush AI Toolkit | Brand visibility in Google AI Overviews | Paid plan | Marketing managers with a mid-size budget |
| Ahrefs (AI Overview tracker) | Branded/unbranded keywords with AI Overviews | Paid plan | SEO managers focused on Google |
| BrightEdge | Brand mentions and citations on ChatGPT | Enterprise | Large companies and agencies |
| Custom Python script + API | Automated queries across multiple AI platforms | One-time development | Companies with an internal tech team |
For most Italian SMEs, the best starting point is the manual test sheet. It takes about 30 minutes a week and produces useful data right away. Paid tools become worthwhile when the number of keywords to monitor exceeds 50-100.
What are the risks of not monitoring my brand's AI visibility?
The main risk is that AI provides wrong or incomplete information about your brand without you knowing. A second risk is that competitors own the category queries while you stay invisible.
AI answers are based on what's published online. If your site has outdated content, an unclear structure or contradictory information, the AI may summarize it inaccurately. As the analysis on the AI citation audit shows, companies often discover that the AI cites competitors or third-party sources in their place.
One concrete figure: between December 2025 and March 2026, company and product sites rose from 55% to 63% of the sources ChatGPT cited for branded queries, according to Built In. Those who invested in structured content gained ground. Those who didn't gave up space.
Why it's a hot topic now: is the AI market in Italy really ready for this kind of analysis?
With 40% of the Italian digital population using AI assistants and over 20 million people exposed to Google's AI Overviews, the channel has reached a critical mass that makes monitoring no longer optional but necessary.
Growth has been fast: in less than two years, AI assistant penetration in Italy has doubled. It's not just young people or tech profiles: it includes decision makers, purchasing managers and professionals looking for suppliers and solutions.
As we discussed in our article on social media and AI Overviews, visibility on these channels is becoming a real competitive factor, not a future trend.
What mistakes are most common when you start monitoring AI visibility?
The most common mistake is testing queries just once and drawing definitive conclusions. AI answers vary by session, by platform and over time: you need repeated, systematic testing.
| Common mistake | Why it is a problem | How to avoid it |
|---|---|---|
| Testing in personalized sessions | AI tailors answers to the user's history, skewing the results | Always use anonymous sessions or new private windows |
| Tracking only ChatGPT | Gemini and Perplexity use different logic and have different audiences | Test at least three platforms in parallel |
| Confusing a mention with a citation that has a link | They are different metrics with different effects on user behavior | Record the two types separately |
| Not distinguishing branded from unbranded | The problems and solutions are different for each type | Keep separate lists and analyze the gaps between the two |
| Updating content without rechecking | You don't know whether your changes affected the AI | Retest the same queries after every significant change |
How do I know if my brand's AI visibility is improving?
The indicators to track are how often you appear in answers, your position relative to competitors, the tone of the answers and the percentage of queries in which the brand is cited with a link.
| Indicator | How it is measured | Positive signal | Check frequency |
|---|---|---|---|
| Branded appearance rate | % of branded queries where the brand appears | Steady week-over-week increase | Weekly |
| Unbranded appearance rate | % of category queries where the brand is cited | Appearing among the top 3 options cited | Weekly |
| Tone of branded answers | Qualitative analysis: positive / neutral / negative | Mostly neutral or positive tone | Monthly |
| Citations with links vs mentions only | Separate count for each type of citation | More citations with links | Monthly |
| Relative position vs competitors | Comparison between your brand and 2-3 direct competitors on the same queries | Brand cited first or more often | Monthly |
So, is it worth starting branded vs. unbranded AI monitoring for your company?
Worth it if your brand already has some recognition in your industry, you operate in a category with comparative searches (software, B2B services, e-commerce, consulting) and you want to know what's being said about you on ChatGPT and Google AI Overviews before your customers find out.
Somewhere in between: you're in a brand growth phase and want to understand whether AI already cites you in category queries. Start with free manual monitoring, then consider more structured tools once you have enough data.
Not yet if you run a purely local, offline business with no significant digital presence. In that case, the AI channel is still marginal for you and it's better to build a base of online content first.
How we see it at SHM
For us, branded vs unbranded monitoring on AI isn't a technical matter: it's about controlling your brand's narrative. AI summarizes what it finds online, and if you haven't built your content in a structured way, someone else, a competitor, a negative review, a third-party source, will fill that gap for you.
We believe technology should give control back to companies, not take it away. Monitoring AI visibility is the first step to understanding where to act: on your site content, on the structure of your service pages, on your presence in authoritative industry sources.
In the SEO for AI projects we manage, the starting point is always a visibility audit that separates branded queries from unbranded ones. That's the only way to tell whether the problem is your AI reputation or your AI awareness, and the solutions are very different. If you want to see how your brand appears today on ChatGPT, Gemini and Perplexity, our strategic AI integration service is the right place to start. If you want to build an in-house monitoring process, we also offer tailored AI training for companies.
Frequently Asked Questions
With a list of 20-30 queries and three platforms to test (ChatGPT, Gemini, Perplexity), you need about 30-45 minutes a week. It takes longer if you want to analyze the tone of the answers or compare your position against competitors.
No. AI answers vary by session, by time of day and over time as models are updated. That's why it's important to test in anonymous sessions and repeat the monitoring regularly, rather than relying on a single test.
The most effective approach is to update and structure the content on your official website, especially the 'About us' and 'Services' pages and the FAQs. AI models pick up new content as it gets indexed. In some cases, OpenAI also offers a channel for reporting incorrect information, but update times are not guaranteed.
It depends on your goal. Branded queries are about reputation: people who already know you are looking for confirmation. Unbranded queries are about acquisition: people who don't know you are discovering the options available. Both matter, but for a growing SME, unbranded queries often have a bigger commercial impact.
Yes. The most accessible method is a Google Sheets spreadsheet with a list of queries to test manually every week. Some Ahrefs and Semrush features also offer limited free versions for monitoring Google AI Overviews. For ChatGPT and Perplexity, manual testing remains the most reliable free method.
Timelines vary. For Google AI Overviews, improvements can show up 2-4 weeks after your site is re-indexed. For ChatGPT and other AI assistants, timing depends on model updates, which happen at irregular intervals. Monthly monitoring lets you catch significant changes.
The question isn't whether AI talks about your brand, but what it says — and are you listening?
Related articles
Discover more articles exploring similar topics, selected to offer you a more complete and stimulating perspective. Each piece of content is carefully chosen to enrich your experience.