Contents:
Our Take: Not Worth It in 2026
Why AI Visibility is Mostly Myth
What to Know Before You Buy
How AI Visibility Tools Actually Work
Where to Invest Instead
Takeaway: Make AI Visibility an ROI Decision
We often get asked for our take on AI visibility tools. These are the SaaS platforms that claim to show how much exposure a brand has within AI conversations. They’re often billed as must-have solutions for GEO / AEO / AI SEO.
While we’re constantly reevaluating our stance, so far it hasn’t changed:
If you already have access to a tool, or money is no object, try experimenting. But if you’re being asked to pay a meaningful amount for the tool, we’re not convinced it’s worth the cost.
We work with many B2B clients, and we know what you want to get out of these tools: visibility into the black box of AI conversations. What is your target audience talking about with AI? How does that conversation influence them toward a certain product category? What are the specifics they ask about and the pain points they reveal? You also want quantifiable metrics you can drop in a reporting deck, tie to KPIs, and use to manage your marketing budget effectively.
But AI visibility tools don’t really provide those answers (even if their messaging suggests they do). What they really offer is more simulation than reality.
The tools aren’t without value, but it’s easy to misinterpret the data they provide and to overstate the actual value of it. While the findings could help inform a strategy, they certainly aren’t ready to be a formal tracking source for KPIs that impact the livelihoods of real people.
Most B2B companies don’t play fast and loose with their marketing budgets. So if you’re wondering whether an AI visibility tool is worth swapping for another line item from your budget wishlist, our answer is a pretty firm no. It’s unlikely the value you’ll derive from these tools is greater than what you’d get from putting the same money towards ads, content creation, site improvements, or market research.
The biggest problem is that these tools can’t really give you the visibility they claim to deliver. The data you see is generally synthetic and represents only one possible answer that an LLM might provide.
So the answer these tools provide is not: Here’s where and how you appear in AI search.
It’s more like: Here’s a guess about what people might be asking LLMs and one possible answer that an LLM could provide if a user were to ask that same question.
Traditional SEO tools can draw from publicly observable search data made available by Google. But discussions with LLMs are not accessible in the same way. Individual users can choose to share conversation links directly with others, but these are generally not indexable by search engines (though ChatGPT briefly allowed this in 2025). While the companies that own the LLMs may have access to this data, they do not share it with third-party AI visibility tools.
I see B2B stakeholders interpret the prompts shown by these tools as fact, as real questions users in their audience asked and actual queries that brought users to their site (or a competitor’s site).
But in reality, visibility tools are showing a simulation, not real, collected data comparable to what you get through something like Google Analytics.
To complicate matters, chatbots are inherently probabilistic. Even if asked the exact same question twice, an LLM will return different answers based on different sources between runs.
It's also important to remember that users are not always querying an LLM one question at a time in a new chat. Variations in prompt wording, conversational context, and saved memory about the user compound the effect of a probabilistic model.
Tools could work around this to improve data; for example, running the same prompt multiple times on the same LLM per day could better triangulate the reality of what a user sees. However, this method would be expensive, which means most tools don’t do it. An AI visibility tool could potentially track different versions of the same general prompt, but not exhaustively, and it would quickly become too expensive to do. Either way, these would still not account for the personal context most users bring to their chats.
That means results shown by typical AI visibility tools are often only a snapshot of one possible response to a much cleaner, clinical prompt than your users are probably asking.
That isn’t without value, but the results you see are much more nuanced than the typical (non-SEO) B2B marketer expects.
While we don’t think AI visibility tools are worth a significant investment, they aren’t useless; they just aren’t as valuable as we want them to be for the price tag.
Assuming you have budget to spare, already paid for a tool, or have free access to a tool, there are a few practical benefits.
In our opinion, the main advantage of these tools is convenience. If you want to track whether you appear in AI responses for multiple prompts, managing that via a tool is faster and easier than manually testing and recording your own data, especially if setting up your own automation process isn’t feasible. The tool will also probably pull helpful data points for you: which competitor brands appear repeatedly, which pages from your site are referenced frequently, and where potential content gaps lie.
If you aren’t SEO-savvy, you might find some value on the research side, too; the prompts you see will have clear user intent baked in from the start, while traditional keyword research can require more expertise and manual analysis of the SERPs (search engine results pages) to pinpoint exact intent. For quick topic ideation, this may be a decent starting place before a writer does deeper research.
Above all, these tools are helpful if you’re interested in watching trends over time. You can get an idea of your relative visibility movement and tweak your strategy accordingly, then track again. If you want a long-term, directional picture of your AI presence, and don’t expect to see actual market visibility, this is a good approach.
AI visibility tools are made by marketers, so they’re great at selling themselves. You may not realize exactly what you’re getting unless you understand how the system works.
Here are the main things we make sure our clients know when they’re considering an AI presence tracker:
The prompts they show you are usually not real user inputs. Tools generally extrapolate prompts from traditional keyword databases. Some may enhance this with a sample set of real data, but most do not document taking this extra step.
Any ranking data presented in the tool is misleading. Rankings don’t exist in AI responses, so attempts to quantify this are either ported over from traditional Google rankings or conflated with position in citation lists that most users will never see (and thus have no impact on your practical visibility).
Most AI visibility data is simulated rather than observed. Typical tracking tools tell you what an AI system could answer when they test a prompt, not what users actually saw. Some newer data sources, such as Microsoft Clarity’s grounding-query and citation reporting, provide observed retrieval activity, but even these do not reveal the original user prompt or the full conversation.
They may imply stability that doesn’t exist. LLM outputs are variable by nature. Running the same prompt again can produce different brands, sources, citations, and wording. Most tools don’t run prompts repeatedly at the scale you’d want to properly triangulate the average response.
Metrics are proprietary and hard to validate. Most companies keep their methodology vague. It may be hard to pin down where the data comes from; be prepared to ask questions and look through documentation.
Your existing tool stack might already have what you need. Many of the same insights can be drawn from some combination of traditional keyword research tools like basic Semrush, brand sentiment trackers like Meltwater, and LLM integrations and workflows through a service like Zapier.
They can create false confidence. The illusion of visibility can be more dangerous than accepting that you can’t really know what’s happening. Dashboards make probabilistic data look definitive. That can misrepresent reality—and the impact of your GEO efforts—to your stakeholders.
We go more deeply into the mechanics behind all of this below.
You can think of AI visibility tools as having two arms: research and reporting:
Research: Finding potential prompts users may be asking that relate to your topic space.
Reporting: Tracking where and how your brand appears in LLM responses to user prompts.
Since there isn’t public data to draw from, AI visibility tools usually base their results on traditional SEO keywords and SERP Feature information, such as the queries that trigger AI Overviews in Google. Most prompts in visibility tools are either pulled wholesale from keyword databases (e.g. “zoos with capybaras”) or extrapolated from keywords into natural language queries (e.g. “What zoos can I visit to see capybaras?”).
More sophisticated tools use clickstream data from a sample set of users who have opted in to tracking. This records the actions users take online, including clicks and search queries, and can provide a layer of reality. But it’s important to note that this is still sampled data, which comes with limitations, and most tools don’t incorporate this in the first place.
One important exception to the general methodology we’ve seen so far is Microsoft Clarity. Clarity surfaces grounding queries—what AI systems actually used to retrieve content before generating an answer—along with the pages that were cited. These are still not the user’s original prompts, but they are meaningfully different from the synthetic prompt tracking described here because they reflect observed retrieval and citation activity that users actually experienced.
Tools take generated prompts, or custom prompts you input yourself, and route each query to your chosen LLM, then report back on the response. A good tool will distinguish citation vs. mention—whether your brand is listed in the sources for the response or actually named in the response—but not all tools do.
Ideally, the tool would run the query multiple times—or run it repeatedly over time and aggregate the data—to create an average profile of your brand’s visibility in responses. But it often doesn’t work this way. The query may only be run periodically if you are not explicitly tracking it.
So what’s actually happening when you see your brand listed as cited or mentioned?
In one run of the prompt, the chatbot mentioned this brand (or listed it in its sources).
If we run that prompt again, or change the wording, the wording of the response will differ.
The LLM may or may not reference your brand again.
The model’s conclusion about your brand may or may not meaningfully differ.
If AI visibility tools aren’t worth the investment, where should you put your money instead? Speaking from the organic strategy and SEO perspective, I would much rather have clients work on the things that drive visibility and measure success through proxy metrics (e.g. keyword rankings) and observable, first-party data.
Here’s where we recommend putting your money before investing in AI visibility tools:
Website performance: Make your site run faster, ensure it renders properly for machine readability, and optimize navigation for bots and users. Add schema if your site is weak on it, but focus on Core Web Vitals first.
Identity and message coherence: Update your main site pages, content library, and offsite profiles (e.g. LinkedIn) to ensure you’re presenting an up-to-date, consistent, and relevant message for AI to synthesize and associate with your brand. Prune your content library to focus your entity on your core topics, deepen your remaining content assets, and add new resources that genuinely help users—either directly on your site or via an AI agent.
Measurement stack (if you don’t already have this): Set up first-party data sources like GA4 and Microsoft Clarity to see what’s really happening on your site, integrate a content delivery network (CDN) to track AI bot traffic, and get traditional SEO tools like Google Search Console, Bing Webmaster Tools, and Semrush to track the data that informs AI search.
PR: Build trust through third-party sources with earned media, guest content, interviews, podcasts, partnerships, and distribution on social media.
Right now, a lot of B2B marketers feel like they need to invest in AI visibility tools because we’re all constantly inundated with FOMO marketing campaigns and pressure from above to report progress against clean, measurable KPIs. But if the tools can’t reliably deliver the information those KPIs are supposed to represent, they’re only creating the appearance of measurement—not necessarily good ROI.
These tools can still be helpful for research. But more often, companies want them as a performance measurement system. In that context, they can do real damage, either by misrepresenting the results of your efforts or swaying strategy in the wrong direction.
So before buying, consider the return rather than the panic. Will this tool give leadership the information it wants? Will it help you build a better visibility strategy? Or is the purchase mainly about adding an AI metric to your dashboard?
Report on the metrics you can have real confidence in. Use AI visibility tools when they genuinely improve research or strategy, not just because you need something to report on.
Want GEO/AEO/AI SEO guidance that’s based in reality? See our AI search expertise.